The Salamander Handbook

When we first thought of starting Salamander, the gap we had in mind was simple. Technology companies needed senior operating help — and most of the help on offer was being provided by traditional consultancies whose people had never actually run the kind of business they were advising. The senior operators who could have made the difference were either still inside companies or had exited and gone elsewhere. There was no obvious place for a company that put experienced operators alongside leadership teams on a flexible basis.

Six of us bootstrapped Salamander, each bringing capital alongside experience and a track record in technology. The intent from the start was to use that combination to help other technology companies — and to do it in a way that traditional consulting could not.

We chose to start with earlier-stage technology companies, going to market through the VC funds that backed them. These were the companies most in need of the experience our team carried; they were also the segment in which we could build a reference base without competing head-on with the larger consultancies. Once we had a body of work behind us, we would move up into the larger companies. We were never trying to be a consulting company. We were trying to be the operator-led alternative.

The differentiator was, and remains, the operating experience the team brings. Every senior person at Salamander has run or led inside a technology business before joining. Some have had consulting roles earlier in their careers; they are operators today, and that distinction is the source of most of what makes Salamander useful to clients.

The fractional model emerged from a structural shift in how senior operators want to work. A meaningful number of executives and ex-founders, having exited their companies, want a portfolio life — several companies across the week rather than one full-time. The Salamander model fits that reality, and it gives our clients access to senior operators at a level of seniority they could not have afforded as a permanent hire. Both sides of the equation made sense at the start, and they still do.

Salamander today operates across Asia Pacific and Europe, the Middle East, and Africa, working alongside leadership teams at scale-up and established technology companies, and the private capital sponsors who back them. This handbook exists because the network has grown to the point where it needs a shared description of how we work — what we believe, what we do, and how we expect each other to operate inside the company and alongside our clients. The three values on the cover — execution, impact, accountability — run through everything that follows. The handbook is for the Senior Associates joining now and for those already inside Salamander who want a single reference. It will continue to evolve as the company does.

We are delighted that you are joining the network. What follows is how we work, what we believe, and what we expect of each other. The rest is up to you.

— Bill Padfield, CEO, Salamander Advisory

Salamander Advisory exists to help technology companies scale. We do this through execution-led transformation: not by writing recommendations and walking away, but by working alongside our clients’ leadership teams to deliver measurable outcomes. The shortest accurate description of who we are is the one we use about ourselves — we are Senior Associates.

The companies we work with are scale-up and established B2B technology businesses, from Series A through multi-billion dollar revenue. They share a common pressure: the need to improve profitability, modernise service models, increase recurring revenue, and strengthen execution discipline at a pace the existing organisation cannot easily generate on its own. Our role is to add the Senior Associate capacity these companies need at the exact moment they need it, in the disciplines where they need it most — Finance, Go-To-Market, and Operations.

We engage fractionally. A Salamander Senior Associate is not a full-time consultant deployed onto a six-month project. The model is built around fractional access to Associates who have done the role at scale before. A CFO-as-a-Service engagement gives a client a senior finance Associate inside their leadership team without the cost or the commitment of a permanent hire. The same fractional model runs across the rest of the leadership team where we operate: CRO-as-a-Service for revenue and sales leadership, CMO-as-a-Service for marketing, and COO-as-a-Service for operations, among others. This shape — fractional, senior, embedded — is what makes Salamander work for clients, and it shapes the way you will work on engagements.

Where we sit relative to traditional consulting companies is best understood by what we are not. Our model is not built around junior teams running engagements under partner oversight. We do not pitch frameworks for their own sake. We do not write deliverables that describe the problem at length and then leave the client to implement. We do not bill discovery work that delivers no decision. The discipline we hold ourselves to is the one we wish more of our clients held themselves to: every engagement should produce a clear path from analysis to decision to outcome, and we own all three.

Geographically, Salamander currently operates across Asia Pacific and Europe, Middle East and Africa. We work directly with leadership teams in those regions and, increasingly, with the private capital sponsors who back them.

There are two further things to know about how Salamander operates as a company. First, everyone here is a contractor — Senior Associates and the company’s leadership alike. No one at Salamander is a salaried employee. The company is a network of Senior Associates who choose to work together. Second, everyone helps build it. Salamander does have a business development function, but it is early-stage: a PR partner and a lead-generation partner who support the company’s pipeline. The expectation is still that every Associate hunts for new deals in their own network and conversations, and that effort is backed by a sales incentive scheme. No engagement is run in isolation from the rest of the company. We grow together and we deliver together.

The implication for you, as a new Senior Associate, is that the standard you will be held to is the standard Senior Associates expect of each other. You are joining a company that does not separate strategy from execution, recommendation from accountability, or the person who assesses the problem from the person who helps the client fix it. That integration is the product. The rest of this guide is about how Salamander delivers it: who we are and what we believe (Part 1), how we work (Part 2), the clients we work with (Part 3), what we do in each of our three disciplines (Parts 4 through 6), the analytical toolkit we apply across every engagement (Part 7), how a Salamander engagement is run end-to-end (Part 8), and where to find the internal tools that support the work (Part 9).

A note on the name. People ask often enough that it is worth saying here. A salamander is one of nature’s most adaptable creatures — it evolves in remarkable ways as it grows. In mythology it is a creature of unusual capability, often described as walking unharmed through fire. Both meanings fit the kind of company we are trying to build. The honest origin, though, is that my youngest son suggested the name, and I loved it. The brand grew from there.

The single fact that distinguishes Salamander from a traditional consulting company is also the simplest to state: everyone here, from CEO and founders through to Senior Associate, has run or led inside a technology business before joining. Not advised on one. Run one. Or led a function, a region, a P&L, or a transformation programme inside one. Operating experience at scale is the entry requirement, not a credential we add to a CV or resume.

This changes the way we work in three ways.

It changes how we assess a problem. We recognise the difference between a stated problem and an actual problem because we have lived it ourselves. A CEO who says revenue is the issue is sometimes correct and sometimes describing a downstream symptom of a sales motion that has broken or a product mix that has drifted. A CFO who says the issue is cost discipline is sometimes correct and sometimes describing a forecasting process that has lost its operating cadence. Reading past the brief to what is actually going on is the first thing we do on an engagement.

It changes the recommendations we make. When we suggest a course of action we are suggesting something we have seen work, or fail, in companies like the client’s. Our recommendations are filtered through experience, not just frameworks. That filter is what lets us anticipate where a recommendation will run into resistance — from the team, from the systems, from the calendar — long before it gets to implementation. The result is practical strategy, built for scale: the difference between a recommendation that survives contact with the organisation and one that does not.

It changes the relationship with the client. We talk to operators differently than advisors talk to operators. The conversation runs at peer level — not because we are senior in title, but because we have done the work. CEOs and CFOs we engage with know within the first few meetings whether the person across the table has run a P&L or has read about running one. That recognition is the foundation of trust on a Salamander engagement, and it is one of the reasons clients open up faster with us than with traditional advisors.

The implication for you, as a Senior Associate, is that the standard is set by the peers you are joining. You bring operating experience to Salamander, or you would not be here. Your job is now to apply that experience consistently, across many client situations, on a fractional cadence, with the methods that the rest of this guide describes. The shift is not from operator to advisor. It is from inside one company to across many, working through clients rather than directly.

That is the work. The rest of the guide gets specific about how it is done.

A traditional consulting engagement ends when the deliverable is handed over. The client thanks the team, signs off the final invoice, and is left to implement. If implementation succeeds, the consulting company shares the credit. If it fails, the client owns the failure. The line between recommendation and outcome is the company’s protection.

Salamander does not have that protection. We take on outcomes. That is the line we have drawn for ourselves, and it is the line that defines what kind of advisor we are.

In practice, taking on outcomes shapes every part of an engagement.

It shapes what we sign up to. Before we agree to an engagement, we ask not just what the client wants explored but what the client wants to be true at the end of it. Margin improved by a defined amount. Recurring revenue mix shifted. A new operating cadence installed and running. A managed services model standing on its own feet. An exit-ready set of financials. The outcome is the brief. The analysis exists to serve it.

It shapes what we deliver. Decks, models, and memos are means to an end on a Salamander engagement, not the end itself. We will write the deck and the memo when they are the right tool for the moment — to align a leadership team, to support a board decision, to brief an incoming hire. We do not write them as proof of effort. The actual deliverable is the outcome, and the supporting materials are whatever turned out to be needed to get there.

It shapes how we work with the client team. Working inside the problem rather than outside it — execution-led partnership — is what makes outcome accountability viable. A team that sits across the table from the client cannot reasonably own an outcome. A team that sits alongside the client, with hands on the work, can. That is why the fractional, embedded model exists. It is the structure that makes outcome accountability honest rather than rhetorical.

It shapes how success is measured. Salamander engagements are scoped to a measurable improvement, and the scorecard runs against that measurement. Not against hours billed. Not against deliverables produced. Not against client satisfaction surveys. Against the change in the business. This is the discipline we hold ourselves to, and it is the discipline our clients hire us for.

There is a line, however, and you need to understand where it sits. We take on outcomes that are within our reach: financial discipline a CFO-as-a-Service can install; a sales motion a CRO-as-a-Service can fix; a managed services model a COO-as-a-Service can build. We do not take on outcomes that depend entirely on factors we cannot influence — a market downturn, a competitor’s move, a customer’s procurement cycle, a regulatory change. Where dependencies sit outside our reach, we say so up front and we scope around them.

The implication for you, as a Senior Associate, is that on every engagement you work, you should be able to state — clearly and in one or two sentences — what outcome this engagement is signed up to deliver. If you cannot, the engagement has not been scoped well enough yet, and your job is to bring that question forward to the Associate leading the engagement before the work goes deeper.

Salamander works with B2B technology companies. That is the market, and it sets the boundary of the work we take on. The companies we engage with sell software, technology-enabled services, infrastructure, platforms, and managed services to other businesses, not to consumers.

Within B2B technology, we work across a wide range of company sizes — from Series A through multi-billion dollar revenue. That range is intentional. Salamander Associates have run technology businesses at points right across this curve, and the problems we are equipped to help with — financial discipline, growth, operating cadence, scaling pains, transformation — appear at every stage. What changes across stages is which problems dominate, and which levers actually move the business.

A Series A scale-up is usually working through the move from a founder-led organisation to a leadership team that can run beyond the founders, from one-off sales motions to a repeatable one, and from a small finance function to one that can produce numbers a board and investors can rely on. A scale-up further down the curve is typically working through international expansion, channel design, or the rebuild of Go-To-Market motions that worked at a smaller size but have stopped scaling. An established technology or services company — one already at hundreds of millions or billions in revenue — is more often dealing with growth slowdown, margin pressure, the migration of a project-based business into recurring service models, or the operating model changes that follow a strategic decision the board has already taken.

We work with three client archetypes inside this market.

The first is scale-up technology companies — later-stage businesses building operating discipline and predictable performance. The engagement need is usually to install the rigour an investor or future acquirer will expect, in a way that does not slow the company down.

The second is established technology and services companies. These are companies facing growth slowdown, margin pressure, or the change in business model that follows a structural shift in their market. The engagement need is usually to find and unlock the next phase of performance without disrupting what is already working.

The third is private capital and boards. These are not the operating businesses themselves but the people behind them — private equity sponsors, venture investors, and the boards they sit on. The engagement need is usually value creation execution, transformation oversight, or transaction readiness on behalf of a portfolio company.

Across every stage of company we work with, one question has moved from optional to central over the last two years: where and how to adopt AI inside the business. For most of our clients this is not primarily a technology decision but an operating decision — one that touches product, Go-To-Market, finance, the cost structure of delivery, and the shape of the team. Almost every client we engage with is working through it in real time, and our engagements increasingly carry an AI dimension whether or not the brief surfaced it explicitly at the start.

Geographically, we operate across Asia Pacific and Europe, the Middle East and Africa. The team is based in those regions, the leadership we work with is in those regions, and the operating rhythms our cadence runs on are theirs. We do not currently operate in North America. If a client conversation surfaces an opportunity in a geography or segment outside this scope, raise it with the Associate leading the engagement rather than declining outright — Salamander’s footprint may evolve, and you will often be closer to the ground than the rest of the company.

The implication for you, as a Senior Associate, is that the first thing to verify on any opportunity is whether it sits inside this perimeter: B2B technology, Series A through multi-billion, APAC or EMEA, and matching one of the three client archetypes. If any of those is unclear, the conversation needs to happen before the engagement is shaped, not after.

The Senior Associate role at Salamander is, at its simplest, a fractional senior practitioner — someone with deep operating experience in technology businesses, applying it across Salamander engagements. The shape of the role varies considerably from person to person. Some Senior Associates work primarily through Salamander; others run their own independent practice and pick up Salamander engagements alongside it. Some carry a single client at a time; others run two or three concurrently. The model is built to flex, because the Associates who join Salamander rarely arrive in identical situations, and the company is more useful when it works around your shape than when it tries to make your shape match its own.

What is consistent is how Senior Associates work with each other and with clients. There are three things worth describing.

First, what each Senior Associate brings to the engagement. On any given piece of work, the integrity of what you produce — the analysis, the recommendations, the model, the storyline, the conversations with the client team — sits with you. So does the relationship you build with the client people you work alongside, and the experience those people have of Salamander through you. We work this way because Associates are senior practitioners and the work does not need to be handed up the chain for cleanup. That trust is one of the things that makes Salamander Salamander.

Second, how engagements come together. Most Salamander engagements are shaped, scoped, and led by another Salamander Associate — sometimes an Associate in a CFO-as-a-Service, CRO-as-a-Service, or equivalent seat; sometimes a fellow Senior Associate who holds the primary client relationship. When you join one of those engagements, you are contributing to the work that Associate is running. You will bring depth on specific threads and carry significant pieces of the analytical and execution work, and the Associate leading the engagement will rely on you to keep them in the loop on how it is landing. Less about reporting; more about the rhythm of working as a pair, or a small team, at pace on someone else’s business.

Third, the relationship between Associates. You are joining a peer network of senior practitioners, and the working relationship reflects that. There is no partner oversight in the consulting sense — Associates work alongside each other rather than below each other. The Associate leading any given engagement is a colleague applying their own experience to the same problem. You will frequently catch things they miss, and they will frequently catch things you miss. The mode is mutual challenge — open, direct, and constructive — because the engagement deserves it, and because the client is paying for the combined judgement of both of you, not the judgement of either alone.

Two company-level norms sit alongside the engagement work, and they are best read as how the network sustains itself rather than as obligations imposed by the company.

The first is helping build the company. Salamander has a small early-stage business development function — a PR partner and a lead-generation partner who support pipeline — but the company still grows mostly through conversations Associates are already having. Bringing deals forward is part of the deal, but the rhythm is yours to set. Most Associates find that an hour or two a week of conscious hunting is enough to make a difference, and Salamander runs a sales incentive scheme so that when you bring a deal in, you share in the upside.

The second is supporting other Associates. The fractional model works because the network flexes — a second pair of eyes when someone needs one, a specific piece of expertise when a colleague is short, a few hours of capacity at a tight moment. Most Associates find that they ask for that kind of help about as often as they give it, and the company runs better the more freely both flow.

Practically, the shape of your week will not look like a full-time employee’s — because you are not one — and it will not look like a traditional consultant’s either. You will move between Salamander client engagements, your own independent work if you have it, internal company rhythms, and the conversations that fill your pipeline. The cadence is yours to manage. For most Associates, that flexibility is one of the reasons Salamander is the right fit; for a few it takes some adjustment in the first few months, and the rest of the network is there to help when it does.

Three words sit above the way Salamander works: Execution. Impact. Accountability. They are the umbrella under which our five values live, and they describe what we are signing up to every time we agree to an engagement.

The five values themselves are:

• Integrity, transparency, and accountability
• Commercial impact over theory
• Execution-led partnership
• Practical strategy, built for scale
• Responsible, scalable growth

The next five chapters take each value in turn. This one introduces the set — why they exist, how they sit together, and how they show up in the day-to-day of an engagement.

The values exist because Salamander is a network of Senior Associates working with each other and with clients largely on trust. There is no consulting-company hierarchy enforcing standards from above; there are no salaried colleagues whose performance can be managed through annual review. The discipline that holds the work together is the discipline each Associate brings to it, and the way that discipline is shared is through values that are written down, lived, and held in common. The values are not a wall poster — they are how the network actually operates.

Each value plays a different role, and they reinforce each other rather than sit in a strict order. Integrity, transparency, and accountability shape how honestly we work — with clients, with each other, and with ourselves. Commercial impact over theory keeps the focus on what changed in the business rather than what was said in the deck. Execution-led partnership describes the posture we take inside an engagement — hands on the work, alongside the client rather than across the table. Practical strategy, built for scale, is the shape of what we recommend — grounded in operational reality, designed to be implemented and to last. Responsible, scalable growth is the kind of growth our clients are trying to build, and the kind of approach we are trying to model in our own company. The values overlap; on most engagements you will draw on more than one at a time.

How do the values show up day-to-day? Rarely by being invoked in conversation. Far more often by being visible in what an Associate does. They are visible in the questions asked in the first meeting on an engagement — the ones that test what is actually true. They are visible in what is left out of a deck — the slides we do not write because they would only describe the problem rather than move the work. They are visible in how a hard message is delivered to a CEO or CFO who would rather not hear it. They are visible in what Salamander walks away from when an engagement is not the right fit. And they are visible in how Associates treat each other inside the network — the willingness to challenge each other, to step in for each other, and to share both credit and blame.

For you as a new Senior Associate, the values work best as a self-check rather than a set of rules. Ahead of any meaningful decision on an engagement — what to recommend, what to leave out, what to escalate, what to push back on — running the decision past the five values is a fast way to see whether you are about to do something Salamander would recognise. If the answer is no, the question is whether the decision is wrong, or whether one of the values applies less strongly in this case. Either answer is fine. What is not fine is making the decision without having asked.

The chapters that follow set out each value in its own right.

Integrity, transparency, and accountability is the value that anchors how Salamander works at the most fundamental level. We operate with absolute integrity and clarity in every engagement. We communicate openly and challenge constructively. And we take ownership of outcomes — not just recommendations.

In day-to-day practice, this value shows up in three places.

The first is in how we talk to clients. Integrity is what is in our recommendations; clarity is how they land. Clients hire Salamander partly because we tell them what we actually think, not what they prefer to hear. The discipline is to do this in a way that is useful — calibrated to the moment, not blunt for its own sake — but the underlying commitment is the same: a CEO or CFO who has paid for Salamander deserves an honest read on their business. That sometimes means delivering a hard message early in an engagement, before the relationship is well-formed enough to make the message easy. The relationship gets built by the honesty, not in spite of it.

There is a real calibration here, and we have learned it the hard way. Salamander once lost a client in its early days because one of us was too direct — treating the client more like a team member they were running than a leader they were working alongside. The transition from running an organisation to advising one, even from a fractional leadership seat, is a real one. The people on the client side are not on your team; they are the team. You are alongside them, contributing your operating experience to decisions they will ultimately make. The honesty stays the same. The register has to shift.

The second is in how we work with each other inside Salamander. We are a network of senior practitioners, and the network only adds value when Associates are willing to challenge each other on the substance of the work. Constructive challenge is part of how we are useful as a pair on an engagement, and it is part of how we improve as a company. The instinct is direct, not deferential: if you see something off in a colleague’s framing, a model, a recommendation — you raise it. The receiving Associate is expected to take it on its merits, not as a critique of their judgement. Even for senior people, this takes some practice — particularly for Associates coming from larger organisations where challenge has typically been routed through hierarchy. At Salamander, the same direct-but-respectful exchange applies regardless of who is challenging whom.

The third is in how we account for what we have done. Accountability at Salamander is not abstract. It means owning outcomes, as Chapter 3 set out. It also means owning mistakes — saying when something we recommended did not work, when something we modelled turned out to be wrong, or when a piece of work we delivered did not land the way we intended. The instinct to retro-fit a narrative that makes us look better than we were is something we deliberately work against. Clients can see it; colleagues can see it; the network is less useful every time it happens. The cleaner alternative is to surface what went wrong, fast, and to do something about it.

Transparency is the connective tissue across all three. With clients, it means being clear on what we know, what we are still working out, and where our judgement could be wrong. With colleagues, it means saying what you actually think rather than the version you have polished for an audience. With yourself, it means being honest about the work you are doing, the time you are putting in, and the standard you are holding yourself to.

For you as a Senior Associate, this value is less about new behaviour and more about consistency. You did not get to where you are by avoiding hard conversations or filing fictional retrospectives. The shift is that you are now doing it in the open, across many engagements, alongside other senior people who hold themselves to the same standard. That standard is one of the things that makes Salamander recognisable to clients, and it is one of the reasons Associates choose to stay in the network.

This value is simple to state: we prioritise results over frameworks. Every Salamander engagement is grounded in delivering measurable improvements — revenue growth, margin expansion, operational efficiency, scalable performance. Strategy only has value when it translates into tangible outcomes.

That may sound obvious. In practice it is a discipline that has to be defended, because every advisory engagement faces pressure in the opposite direction. The pressure comes from inside the engagement (it is easier to produce a framework than to produce a result) and from outside it (clients sometimes ask for a framework because it feels like progress). Holding the line on commercial impact is one of the things that distinguishes how Salamander works, and it is one of the most important habits a Senior Associate will form.

In practice, commercial impact over theory shows up in three ways.

It shows up in how we scope engagements. The first questions in a scoping conversation are not “what should we analyse” but “what does the business need to be true at the end of this work, and how will we know if it is.” If the answer is a number the client can point to on a board pack — margin up, churn down, sales productivity improved, cash runway extended — the engagement has a frame that can hold it. If the answer is “a strategy,” “a recommendation,” or “an analysis,” the engagement is under-scoped, and the conversation has to keep going until the commercial impact becomes specific.

It shows up in what we bring to the work. We use frameworks, but we use them as scaffolding for analysis, not as deliverables in themselves. A two-by-two matrix that helps a CEO see two trade-offs cleanly is useful. The same matrix in a slide titled “strategic framework” with no business decision attached to it is not. The test is always the same: does this move the business, or does it only describe the business? The frameworks we keep are the ones that pass the test.

It shows up in how we write up our work. Salamander deliverables read like internal memos written by people who have run the business, not consulting decks produced from the outside. Headlines say what we found and what to do, not what we analysed. Models calculate the actual financial impact of the change being recommended, not a generic illustration of how the model would work. Slides exist to support a decision, not to demonstrate effort. When we write something that would not be useful to someone running the business, we cut it.

That does not mean Salamander avoids decks. Over time, we have built up a toolkit of a substantial number of PowerPoint decks covering recurring scenarios — the kinds of conversations and issues that come up again and again across our clients. Associates draw on the toolkit whenever doing so accelerates the work. A deck from the toolkit can be the fastest way to open a discussion with a client team or to frame an issue the client is already wrestling with, and not using it when it would help is a wasted advantage. What we do not do is reach for a deck because it is convenient, or substitute a deck for a clearer answer. The test is the same one we apply everywhere: does this move the engagement toward an outcome, or does it only fill time.

There is an AI corollary worth naming. Clients increasingly ask Salamander for help on AI, and the temptation in that work is to substitute frameworks (“the AI adoption journey,” “a maturity model”) for actual commercial decisions. We hold the same line on AI as we do on every other topic. AI as a recommendation is empty. AI as a measurable change in cost, in cycle time, in revenue per head, or in the shape of a customer-facing function — that is commercial impact, and that is what we sign up to help with.

A related responsibility sits with every Associate. The pace of change in AI is faster than the pace of change in any other tool in the operator’s toolkit, and the half-life of confident assumptions about what it can and cannot do is short. Every Associate at Salamander — regardless of discipline — has a personal responsibility to keep up with how AI is being applied in B2B technology businesses: in finance functions, in go-to-market motions, in delivery and managed services, in the products our clients sell. That does not mean becoming an AI specialist. It means reading, experimenting, and staying close enough to the practice that you can tell a client what is real and what is hype. A Salamander Associate who is six months behind on AI is a Salamander Associate carrying a real gap into a client conversation. The expectation is that we close that gap continuously.

For you as a Senior Associate, the practical test on any piece of work you produce is the one this chapter keeps circling: does this move the business, or does it only describe the business? It is the test the Associate leading the engagement is using on your work, and it is the test you will get faster and more confident at applying to your own. Over time, most Associates find that they cannot easily produce work that fails it — and they cannot easily consume work that fails it either. That is the muscle this value builds.

This value describes the posture we take inside an engagement. We work inside the problem, not outside it. Salamander operates as an extension of the client’s leadership team — hands-on, pragmatic, and committed to seeing initiatives through to completion. We are doers, not observers.

This is the value that turns Salamander’s stated commitment to outcomes (Chapter 3) into something the client actually experiences. Without this posture, outcome accountability is a slogan. With it, the way the client experiences the engagement day-to-day is the way an internal hire would feel — only with someone who has done this before.

It shows up in several practical ways.

We sit alongside the team, not above it. A Salamander Associate on an engagement attends the client’s operating reviews, joins the relevant working sessions, reads the same data the team is reading, and gets involved in decisions in the rhythm they are being made — not at a remove. The point is to be in the same problem space as the people whose job it is to solve it, and to feel the same constraints they feel. That is what produces recommendations that survive the day after they are made.

We work iteratively, not in waterfall. A traditional consulting model designs a recommendation in isolation, presents it once it is complete, and then steps back. Execution-led partnership runs the other way around. We try things, learn from them, adjust, and re-try. Much of the work on a Salamander engagement is iterative — a model that gets sharper through three or four versions; a sales motion that gets tested before it is rolled out broadly; a managed services model that runs as a pilot before it runs as a programme. The discipline is to plan for iteration up front, and to write recommendations that can absorb new information without falling apart.

We are pragmatic about what is actually possible. The version of a recommendation that gets implemented is almost never the version that would be ideal in a vacuum. It is the version the team can actually execute, with the resources, the systems, and the calendar they have. Salamander engagements are designed around what is implementable, not what is theoretically tidy. Getting the right thing done is better than designing the perfect thing — and getting it done at the right altitude, for the team and the calendar that actually exist, is harder than it sounds.

We stay engaged through delivery. The fractional model lets us stay alongside the client past the point where a project-based engagement would have ended. When a recommendation runs into an obstacle, we are still there. When a programme stalls, we are part of restarting it. The commitment is to the outcome, not to the duration of a phase. That is one of the practical differences between a fractional CFO-as-a-Service and a one-off finance review, and it applies across all of Salamander’s disciplines.

The posture has a real cost, and it is worth naming. Working inside the problem means you cannot keep the comfortable distance an external advisor enjoys. You will be present when things go wrong, not just when things are being designed. You will sometimes be associated with parts of the work that did not land, because you were close enough to them to share the outcome. That association is what makes the upside genuine, and it is why this value pairs so closely with integrity and accountability. The two stand or fall together.

For you as a Senior Associate, the practical implication is that you should expect to be in the work, not above it. The chapters in Parts 4 through 6 set out what that looks like in each of Salamander’s three disciplines. Part 8 sets out the day-to-day cadence of running a Salamander engagement. This chapter is the posture they sit inside.

This value describes the shape of what Salamander recommends. We combine strategic clarity with real-world execution experience. Our recommendations are grounded in operational reality — designed to be implemented, scaled, and sustained. We anticipate challenges not from theory but from experience in building and transforming businesses. We bring structure, discipline, and focus to complex environments.

The chapter on commercial impact (Chapter 8) made the case that we judge our work by whether it moves the business. This chapter is the answer to the obvious next question: what kind of recommendation is most likely to move the business, and what kind tends to stall in the implementation phase. Most Salamander Associates arrive with a clear instinct about this, because most have been on the receiving end of recommendations that stalled. The instinct gets sharpened and gets applied across many engagements, but it does not have to be acquired from scratch.

A few things make a Salamander recommendation recognisable.

It is anchored in operational reality. The shape of the team that has to execute, the systems they have to work with, the calendar they are running on, and the bandwidth they have on top of their day jobs — these are inputs to the recommendation, not afterthoughts. A recommendation that does not account for them is not “ambitious”; it is a recommendation that will not happen.

It anticipates resistance from experience. Most recommendations encounter the same kinds of resistance at the implementation stage: a senior leader who quietly disagrees and slow-rolls the rollout; a system change that needs more time than the plan assumed; a stakeholder whose support is needed and was not consulted early. Salamander Associates have seen these patterns before and shape the recommendation to absorb them. That is what “anticipating challenges from experience” actually means in practice.

It is built for the scale the business is heading toward, not the scale it is at today. A recommendation that solves the problem for the size the company is now, but breaks at the size it will be in twelve or eighteen months, is a partial recommendation. Where it is sensible to do so, we design with the next inflection point in mind: a finance function that can produce numbers at the next scale; a sales motion that survives a move from twenty salespeople to fifty; a managed services model that can absorb new clients without re-architecting itself. The judgement of how far ahead to design is itself part of the recommendation.

It is structured for sustainment, not just launch. Most transformations fail not in the design or in the launch, but in the year after. Salamander recommendations are written so the practices we install continue after we step away. Operating cadences, KPIs, controls, governance structures, and the discipline of running them are part of the deliverable, not addenda. Sustainment is a design problem, and we treat it as one.

It carries structure, discipline, and focus into complex environments. Our clients’ situations are usually messy — multiple programmes running, several pressures at once, leadership teams with limited room to think. Part of what Salamander brings is the ability to hold structure in that mess. A clear set of priorities. A small number of outcomes that matter most. A workplan that can be communicated to a board in five slides. Bringing this kind of focus is not a soft skill; it is the difference between recommendations that get attention and recommendations that get implemented.

For you as a Senior Associate, the bar is to produce recommendations that pass all five of these tests in combination. It is rare to fail any one of them on its own. It is much more common for a recommendation to be strong on three or four of them and weak on one — and that is usually where the recommendation stalls when implementation begins. A useful habit is to check each draft of a recommendation against all five before sharing it with the Associate leading the engagement.

This value describes the kind of growth Salamander helps clients build. We help businesses grow in a way that is sustainable, disciplined, and aligned to long-term value creation. It is the destination most of our engagements are pointed at — and it is the way we try to grow Salamander itself.

There is a reason this value sits inside our values document rather than being assumed. The technology industry has a long history of celebrating growth without distinguishing between growth that compounds and growth that consumes itself. Customers acquired faster than they can be served. Revenue booked faster than it can be recognised. Headcount added faster than management can absorb. None of these is growth in the sense that creates long-term value. Each of them is a leading indicator of a problem that becomes acute a year or so later. Salamander engagements are designed to help clients build the other kind.

Three components run through what “responsible, scalable growth” actually means.

The first is building repeatable revenue models. A business that grows through one-off wins is not yet scaling — it is selling, in a way that depends on every deal being engineered individually. The shift to a repeatable revenue model is one of the most consequential a B2B technology company can make, and it shows up across all three of Salamander’s disciplines: in Finance, as the move toward recurring revenue and accurate revenue recognition; in Go-To-Market, as the build of a sales motion that produces predictable pipeline rather than depending on heroics; in Operations, as the design of delivery models that absorb new customers without re-architecting the company. Repeatability is what makes growth predictable, and predictability is what makes a company worth more.

The second is strengthening financial and operating discipline. Growth that outruns the company’s ability to control it is not growth — it is risk. Salamander’s CFO-as-a-Service work, the operating cadences we install, the KPIs and controls we build, the reporting standards we hold ourselves and the client to: these are not separate from the growth agenda. They are how growth becomes durable. A company that can grow at thirty percent a year but cannot tell its board with confidence what its margin profile will be in six months has a discipline problem, not a growth problem.

The third is creating organisations that scale beyond individuals. Founders and senior leaders are usually the engines of growth in the early stages of a technology business — and the bottleneck in the later stages. A company that depends on a small number of individuals to make every important decision cannot scale past those individuals’ capacity. The shift is to a leadership team that can run beyond the founders, an operating cadence that produces decisions reliably rather than heroically, and an organisational structure that does not need the CEO in every conversation. Building this is slow, sometimes uncomfortable, and almost always under-prioritised relative to its long-term importance. We make a point of putting it on the table.

The balance the value asks us to hold is between ambition and control. Ambition without control produces growth that does not last. Control without ambition produces a business that survives but does not move. Salamander is not in the business of recommending either extreme. The work is to design growth plans that stretch the company but not the rails, and to install the discipline that turns ambition into compounding rather than thrashing.

This applies to Salamander too. We are growing as a company — adding Associates, opening new conversations in our regions, building out the toolkit, deepening our presence with private capital sponsors. The same discipline applies. We try to add Associates the network can support; to take on engagements we can deliver against; to grow our own pipeline in a way that does not depend on heroics. If the value is what we ask our clients to live by, we have to live by it ourselves.

For you as a Senior Associate, the practical implication is to recognise this as the destination of most of the work you will do — and to keep it in view even when an immediate engagement is focused on a narrower outcome. A pricing change, a sales motion redesign, a cost programme, a managed services build: each is a step on the longer path that this value describes. The work lands harder when the Associates closest to it can see how it connects to that longer path.

This chapter, and the four that follow, set out how Salamander works. The first principle is the one this chapter is named for: we are execution-led, not deck-led. Part 1 covered the why — outcomes over advice, commercial impact over theory, the posture of working inside the problem rather than outside it. This chapter is about three practical consequences: how an execution mindset reshapes scoping, work product, and pace.

A note before the three. “Execution-led, not deck-led” does not mean we do not produce decks. As Chapter 8 set out, Salamander maintains a substantial toolkit of decks for recurring scenarios, and Associates draw on them whenever they accelerate the work. The distinction is that decks are an instrument in service of execution, not the substance of what we deliver. The same principle applies to models, memos, frameworks, and any other artefact a traditional consulting model might mistake for an outcome.

A related note. “Execution-led” also does not mean we will not develop or document a strategy when a client asks us to. Occasionally a client wants a strategy produced — a total company strategy for a scale-up, a GTM strategy, a strategy for a specific function. The ask is not frequent at Salamander, but it is well within what the network can do, and it sits naturally alongside our discipline-led engagement shapes. The same principle applies: the strategy exists to serve a decision and a set of outcomes, and the engagement is built around landing those, not around producing the document itself.

How execution-led thinking reshapes scoping.

The scoping conversation at the start of a Salamander engagement looks different from the equivalent conversation at a traditional consultancy. Our scoping documents include deliverables and timelines — both Salamander and the client need that clarity, and we set those out clearly. But the emphasis sits on the outcomes the engagement is signed up to produce, and the deliverables and timeline are scoped to serve those outcomes rather than the other way around. The questions that anchor the conversation are about what state the business needs to be in at the end of the work, what decisions need to be made along the way, and what role Salamander needs to play in making them.

A small consequence with a large effect: the Salamander workplan reads top-down from the outcomes. Deliverables and dates live inside it, but they are the path, not the destination. Senior Associates new to Salamander sometimes find this orientation different from what they are used to. It becomes natural quickly.

How execution-led thinking reshapes work product.

If decks are instruments, what we produce on a Salamander engagement is what the engagement actually needs to land. Sometimes that is a deck — to brief a board, align a leadership team, or open a difficult conversation. Sometimes it is a model that resolves an open question. Sometimes it is a memo, a workplan, a draft job specification, an interview guide, or nothing written at all — a sequence of well-prepared conversations that move the client from one position to another. The shape is chosen for the moment. That requires you to be comfortable producing whichever artefact the situation needs, and equally comfortable producing nothing where nothing is needed.

A useful test on any artefact you produce is the question: what decision or change does this enable, and who is the audience whose decision or change matters? If you cannot answer in one sentence, the artefact is probably solving for the wrong thing.

How execution-led thinking reshapes pace.

Execution-led engagements run at the pace of the client’s operating rhythm, not the pace of a consulting workstream. The client has a board cycle, a forecasting cadence, a quarterly close, a customer renewal calendar, a hiring sequence. Salamander Associates are inside those rhythms and have to move with them. That usually means moving faster than a traditional consulting timeline would allow — a recommendation that lands two weeks after the board meeting it was meant to inform has missed the point. It occasionally means moving more slowly, when the right thing to do is to wait for the next operating review or the next data point rather than force a conversation before it can land.

The unifying idea is that pace is set by what the business needs, not by what the engagement plan said in week one. A workplan that survives unchanged for six weeks is usually a workplan that has stopped being useful.

For you as a Senior Associate, the practical implication is that your daily question is not “what deliverable am I working on” but “what is the next thing that needs to happen for this engagement to land, and what is the smallest thing I can produce that makes it happen.” That question is the one this chapter is really about.

If Chapter 12 described how execution-led thinking reshapes the structure of a Salamander engagement, this chapter describes how the engagement moves forward inside that structure. The phrase that captures it is the one in the chapter title: from analysis to decisions to outcomes. The chapter is short because the principle is simple — though the practice is harder than it looks.

Most advisory work has three phases in some form. Analysis: looking at the data, talking to the team, understanding the situation. Decision: a point at which the client commits to a course of action. Outcomes: what happens after the decision, in the form of change in the business.

The Salamander discipline is to move clients through these three phases with as much momentum as the situation can absorb, and to spend the least time possible in the analysis phase consistent with the decision being a sound one. Not because analysis is unimportant — it is the foundation — but because most engagements run aground when analysis stretches longer than the decision it is meant to inform actually requires. Analysis exists to serve a decision; once the decision is in view, the additional analysis has diminishing returns.

Three behaviours follow from this.

We point at the decision early. From the first conversations in an engagement, we are working backwards from the decision the client needs to make. What is the question? Who is the decision-maker? When does it need to be made? What standard of evidence do they need to make it confidently? Holding the decision in the foreground keeps the analysis focused. It also forces a useful kind of honesty — about whether the right decision-maker is in the room, about what evidence will actually move them, and about whether the engagement is set up to reach the decision at all.

We recognise when a decision has actually been made. Decisions in technology businesses often get discussed without being made. A leadership team meets, debates, leans toward an answer, and breaks for the day. The next week, the same conversation happens again. Salamander Associates have to be calibrated to distinguish a decision that has been made — clearly, with the decision-maker visibly committing — from one that has merely been discussed. Where a decision has not actually been made, we say so, and we name what is needed for it to be made. That is sometimes a more uncomfortable contribution than the analysis itself.

We do not stop at the decision. The hardest part of most engagements is what happens after the decision is taken — the period when the change has to be installed in the operating reality of the business. Many advisory engagements wind down at the decision point, on the assumption that implementation is the client’s problem. Salamander engagements are usually structured the other way. The decision is the midpoint of the engagement, not the end. The fractional model exists precisely so that we can stay alongside the client through what comes next — the operating cadence that has to be installed, the team that has to be repurposed, the system change that has to actually happen.

A pattern worth noting. When a Salamander engagement appears to be stuck, it is usually stuck in one of two places: an analysis that is not converging on a decision, or a decision that has been taken but is not being executed against. Recognising which of the two it is takes practice. Assessing it accurately is one of the things the Associate leading the engagement will be looking to you to help with.

For you as a Senior Associate, the practical implication is to keep the three-phase mental model in mind at all times. On any given day, you should be able to answer two questions: what phase is this engagement in, and what would it take to move it to the next phase. If you cannot answer the first question, the engagement is under-scoped. If you cannot answer the second, the engagement is under-driven. The Associate leading the engagement will frequently ask you both — and over time, you will ask them yourself before they have to ask you.

A Salamander engagement looks different from a traditional consulting engagement in one immediately visible way: there is no team of junior consultants beneath the senior people. The Associate working with the client is the Associate doing the work. There is no leverage model behind the scenes.

Engagements vary in shape. Some are run by a single Senior Associate. Others have two Associates working alongside each other — sometimes two Senior Associates, sometimes a Senior Associate working with an Associate in a CFO-as-a-Service or equivalent seat. None of those structures involves a junior layer doing the actual work. That is the consistent piece.

There are consequences to this, and they go in several directions.

The consequences for the client. The client does not have one relationship managed at the partner altitude and a working team they barely meet at another. The person who scoped the engagement is the person reading the data, building the model, drafting the recommendation, and presenting it. The conversation a client has at the start of week three is with the same Salamander Associate who heard the original brief. That continuity is one of the reasons we are able to compress timelines in ways that traditional consulting engagements struggle to match — the people doing the work do not have to be briefed in between conversations, because they were in the conversations.

It also shapes how the engagement is priced. Salamander does not bill by the hour. Engagements are priced as a retainer for an XaaS seat (CFO-as-a-Service, CRO-as-a-Service, and the rest), as a scoped fee against an outcome, or as a combination of the two. The client is paying for senior capacity and for the outcomes that capacity is signed up to produce — not for hours on a timesheet. A Salamander engagement is rarely cheaper in unit terms than a traditional one. It is usually cheaper in total, because there is no machinery between the partner and the work, and because the time spent producing materials that exist only to justify the team’s existence does not exist at all.

The consequences for Salamander. Because the working team is small and senior, our judgement is more concentrated and our analytical bandwidth is more constrained. We choose engagements where the value of senior judgement is high relative to the value of analytical breadth. We are usually not the right answer when a client primarily needs analytical volume — a large benchmarking exercise, a survey-driven research piece, an industry scan that requires a lot of hands. We are usually the right answer when the work requires repeated calibration with the client’s leadership team, when a hard recommendation has to be carried into implementation, or when fractional senior capacity is the missing ingredient inside the company itself.

The consequences for Senior Associates. The most important practical consequence is the one Chapter 5 already touched on: there is nobody beneath you doing the work. You are the team, alongside whichever other Associate is on the engagement. You do not delegate analysis to anyone else. You build the model. You write the memo. You prepare the deck — drawing on the toolkit where useful. You sit through the operating review and take what you heard into the next piece of work. The artefacts and the thinking belong to the same person. It is one of the reasons the role is satisfying for ex-operators, and it is one of the reasons it can be a difficult adjustment for people coming from a leveraged consulting model where the working pieces were always handed down.

A note on workload calibration. Because there is no junior layer, the senior-led model has limits. We cannot absorb unbounded scope on a fractional cadence. The Associate leading an engagement is responsible for ensuring scope and capacity match. If you find yourself on an engagement where the scope keeps expanding into volume work that another model would handle better, raise it. When Salamander engagements run into trouble, it is more often because the scope crept past the size the senior-led model can sustain than because the team was unable to do the work.

For you as a Senior Associate, the implication is to embrace the absence of leverage rather than work around it. Your value is not in how much analysis you can produce; it is in how much of the right analysis you can produce, with judgement attached, fast enough for the client’s operating rhythm. That is a different kind of productive than the leveraged model assumes, and it is the one this chapter is really about.

Fractional engagement is Salamander’s primary delivery model. The chapters before this one have referenced it repeatedly; this chapter sets out what it actually means and why it changes the work.

The simplest description is that a Salamander Associate plugs into a client’s leadership team on a part-time but durable basis, taking on a senior operating role for the period the engagement requires. The most established version of this is CFO-as-a-Service — a Salamander Associate serving as a client’s CFO for two or three days a week, over a period of months or longer, inside the leadership team rather than alongside it as an advisor. The same shape applies across the rest of the C-suite where Salamander operates: CRO-as-a-Service for revenue and sales leadership, CMO-as-a-Service for marketing, and COO-as-a-Service for operations, among other configurations.

Three things distinguish fractional engagement from the alternatives.

It differs from a project-based consulting engagement. A project engagement scopes a question, produces a recommendation, and ends. A fractional engagement scopes a leadership role and stays in it through the period the role is needed. A CFO-as-a-Service engagement does not produce a “finance recommendation” and hand it back; it produces the finance function actually running at the right standard, week after week, while the role is occupied.

It differs from a full-time hire. A full-time CFO is on the client’s payroll, integrated into the company permanently, and represents a multi-year commitment in cost and management overhead. A fractional CFO-as-a-Service brings the same level of seniority for the share of the week the company actually needs it, without the cost or the long-term commitment. For most B2B technology companies between Series A and the point at which a senior leadership team comes fully together, the fractional model is an inflection-point fit: too small for a full-time hire to be justifiable, too important to leave the role empty.

It differs from arm’s-length advisory. A traditional advisor sits outside the team and is consulted when needed. A fractional Salamander Associate is inside the team — attending the operating review, signing off the board pack, having the difficult conversation with the underperforming function head, taking ownership of decisions in their domain. The work is embedded, not external. That is the line that gives fractional engagement its weight, and it is the line that makes outcome accountability honest. Salamander Advisory is, in that sense, advisory of a particular kind: senior advisory delivered from inside the leadership team rather than from across the table.

Several practical features follow from this model.

The duration is set by the role, not by a project plan. A fractional engagement runs for as long as the role is needed at the seniority Salamander brings — sometimes three or six months while a permanent successor is recruited; sometimes two or three years while a company moves through a growth phase that requires sustained senior capacity; sometimes longer where the fractional model is the right answer for the size the company has chosen to operate at.

The cadence inside the engagement is set by the client’s operating rhythm. A fractional CRO-as-a-Service is in the weekly sales pipeline review, the monthly QBR, the quarterly board cycle, the annual planning conversation — the same rhythms the role would be in if the seat were full-time. Salamander Associates calibrate their own time inside those rhythms, and the rhythms set the pace of the work.

Multi-client capacity is part of the model. A Senior Associate may hold one fractional engagement at the level of a full-week commitment, or two at half-week commitments, or three smaller engagements alongside their own independent practice. The model assumes Associates are senior enough to manage their own capacity across engagements, and the operating discipline that allows them to do so is part of the role.

A build-operate-transfer pattern is one of the recurring fractional shapes. The Associate occupies the seat (CFO-as-a-Service, CRO-as-a-Service, or another), builds the function to the standard it needs to run at, and stays in it long enough to know what the permanent successor needs to look like. When the client is ready to hire, Salamander helps draft the job description, sources candidates through our talent partner Olofsson (www.olofsson.ai), interviews the shortlist alongside the client, and trains the incoming hire through the handover. The Associate exits once the successor is operating at the standard the role requires. This sequence gives the client time to recruit deliberately rather than under pressure, and gives the new hire a function that is already running rather than one to be built from scratch on top of starting in the role.

A note on what fractional is not. It is not “part-time consulting.” It is not “advisor on retainer.” It is a leadership seat in the client’s team, occupied by a senior person, on a less-than-full-time basis. The seat carries the accountability of the role for the period it is occupied. That is the substantive difference, and it is the one that gives the engagement its shape.

For you as a Senior Associate, the practical implication is to think of yourself as inside the client’s leadership team for the period the engagement lasts — with the operational responsibilities and political realities that come with that — rather than as an external party producing pieces of analysis. The work is run from inside, on the client’s cadence, against the role’s standard. The rest of this guide is, in detail, about how to do that well.

The clients we work with are usually under real pressure. A CEO whose board is asking why growth has slowed. A CFO trying to close the books with a forecast that has slipped twice. A founder who is realising that the leadership team they hired two years ago is not the team they need now. A COO whose service margins are eroding and who cannot quite name why. The decisions these people are weighing are consequential, often personally as well as commercially, and they are weighing them while running the business at full speed.

How Salamander Associates work alongside people in this state matters. This chapter is about that posture.

The first thing to acknowledge is that the pressure is real. Some of it is well-founded; some of it is amplified by circumstances; some of it is internal pressure the client is putting on themselves. The composition of the pressure is part of what we are reading in the early weeks of an engagement, because the response that is helpful depends on which kind of pressure we are seeing.

Several things characterise how Salamander Associates show up in this context.

We bring a steadiness the client can borrow against. A leader under pressure who walks into a meeting with their fractional CFO-as-a-Service or COO-as-a-Service has, ideally, walked into the room with someone calmer than they are. Not because the situation is less serious than they thought, but because we have been in situations like this before — usually as operators inside companies of our own. That experience translates into a calmer reading of the moment. The steadiness is not performed. It comes from pattern recognition, and from having lived through analogous situations and emerged on the other side.

We listen first, harder, and longer than most people expect. Leaders under pressure are often surrounded by people offering opinions before they have understood the situation. The most useful contribution in the first weeks of an engagement is usually to listen — to the CEO, to the rest of the leadership team, to the people one or two layers down, to the data — and to refrain from pre-loading recommendations into the conversation. The recommendations land harder later precisely because the listening was done first.

We are honest about the assessment. Chapter 7 set out the honesty value; this chapter is where the value gets tested. A leader under pressure is sometimes hoping the answer is something other than what we can see. The discipline is to deliver the actual assessment — clearly, calibrated to the moment, but without softening past the point of usefulness. As Chapter 7 also acknowledged, the calibration has limits in the other direction: the people on the client side are not on our team, and the register of how an operator might challenge a peer inside their own company is not always the register that fits a fractional seat alongside a client.

We are disciplined about what we say outside the room. Working alongside leadership under pressure means hearing things that are not for general circulation — strategic moves not yet announced, performance issues with specific people, conversations with sponsors that the team does not yet know about. Discretion is not a soft skill in this context; it is a precondition for being trusted with the access that makes the work possible. The information we hold for a client stays held until the client decides to release it.

We pace ourselves around the cadence of the pressure. Some engagements run inside a quiet stretch, where the work moves at the rhythm of operating reviews and planning cycles. Some engagements run inside a peak — a board cycle, a fundraise, an exit, a major customer at risk — where the cadence is daily and the client needs Salamander to be available in the moments the pressure is highest. Reading where in that range a given week sits, and adjusting your availability and tempo accordingly, is part of the role.

A note on personal calibration. Working alongside leaders under pressure can be intense, particularly during the peak periods. Most Associates have done this in their own operating careers and know what sustained pressure feels like from the inside. The fractional model is partly designed to absorb this — the Associate carries the pressure for the share of the week the engagement occupies, and steps back to other engagements, internal Salamander work, or their own time the rest of the week. Make use of that rhythm.

For you as a Senior Associate, the practical implication is to bring two things into every engagement: a clear-eyed reading of the pressure the client is under, and the steadiness to be useful inside it. The work is technically demanding. It is also, much of the time, an exercise in being the most composed person in the room. Both pieces matter.

Scale-up technology companies are the first of three client archetypes Salamander works with. This chapter sets out what a “scale-up” looks like in our context and what kinds of work we are typically called in to do.

A scale-up, in Salamander’s working definition, is a B2B technology company that has moved past the question of whether the business works and is now working through the question of whether it can be made to work at a meaningfully larger size. The earliest scale-ups we work with are roughly Series A — old enough that product-market fit is broadly established, young enough that most of the operating infrastructure of a larger company has not yet been built. The latest are companies several years further on, sometimes already at hundreds of millions in revenue, that are still scaling rapidly enough that they have not yet stabilised into an established operating shape.

What unifies scale-ups across this range is that they are building operating discipline and predictable performance while still moving fast. The two have to coexist. A scale-up that loses its growth in the process of installing discipline has failed at the transition; a scale-up that grows without installing discipline accumulates debt that will be paid in the year or two after. Most of our scale-up engagements sit inside that tension.

A few patterns recur.

The leadership team is being rebuilt. The team that took the company from zero to Series A is rarely the team that takes it from Series A to the next stage. New senior hires are arriving, founders are stepping back into different roles, and the operating cadence is being rewritten around a wider group. Salamander Associates often arrive into this transition. A fractional CFO-as-a-Service or COO-as-a-Service can hold the seat while a permanent hire is recruited, can mentor a newly promoted internal candidate into the role, or can do the work that the existing seat-holder is not yet able to do alone.

The Go-To-Market motion is being rebuilt for repeatability. The sales motion that worked when the founders sold every deal personally rarely survives the first new layer of revenue. The work — and Salamander is regularly engaged in this work — is to redesign the motion for predictability: an ICP that is actually fit, a sales process that does not rely on heroics, channel decisions that work for the company’s economics, and a pipeline cadence that produces forecasts a board can rely on.

The finance function is being upgraded from bookkeeping to investor-grade. Scale-ups frequently arrive at a stage where the numbers being produced are accurate enough to file but not yet structured enough to run the business on. Building the cadence, the controls, the KPIs, the rolling forecast, and the reporting standards that a board, investor, or future acquirer will expect is one of the most common Salamander engagements at this stage.

The cost structure is being held to a discipline the company has not previously needed. Series A and B funding often allowed the company to outrun its cost base; later-stage scale-ups have to bring revenue and cost back into balance — through margin programmes, cost reshaping, or the migration of a project-based revenue mix into a recurring one with better unit economics. Each of those is a recurring engagement type for us.

International growth is being attempted, often for the first time. The home market stops being sufficient; new geographies open. The decisions about which markets, in what sequence, with what entry model, and at what investment level are decisions Salamander is regularly asked to support — and the operational lifts that follow those decisions are usually larger than the client expected.

A note on qualifying scale-up opportunities. The patterns above describe the work scale-ups need; they do not on their own describe an engagement Salamander should sign up to. A scale-up that recognises the value of senior fractional capacity but is not in a position to pay for it is not yet a client. We have spent time, in the past, on early-stage opportunities that wanted what Salamander brings without the budget to engage us properly, and that time has not produced outcomes for either side. The discipline at the qualification stage is to confirm three things alongside the obvious “is there a problem we can help with”: is there a real decision-maker, is there a real budget, and is the company prepared to pay for the seniority and the outcomes Salamander is signing up to deliver. A “yes” on the first and “not yet” on the second and third is a relationship to maintain rather than an engagement to scope. Be honest about the difference, early.

How to read a scale-up in the early weeks of an engagement. The signals worth attending to are not the headline growth number but the pattern beneath it: the shape of the pipeline, the durability of the customer base, the predictability of the finance function, the depth of the leadership team. A scale-up that looks healthy on the top line and unhealthy on these structural indicators is the scale-up most likely to need Salamander’s help — and it is often the one most at risk of denying it does.

For you as a Senior Associate, the practical implication is to expect a degree of organisational churn alongside the work itself. Scale-up engagements rarely run in a stable environment. New hires are joining, structures are being adjusted, conversations with investors are running in parallel. Reading the scale-up well — what is being built, what is being repaired, what is being avoided — is one of the things this chapter is really about, and it is the foundation for everything that follows.

Established technology and services companies are the second of three client archetypes Salamander works with. They are companies that have moved past the scale-up phase, settled into an operating shape that works at their current size, and are now facing a different category of problems: growth that has slowed, margin under pressure, or an operating model that needs to change in response to a structural shift in their market.

The size range begins somewhere above the scale-up phase — companies with hundreds of millions of dollars in revenue, established leadership teams, operating infrastructure that has been refined over years — and runs up to multi-billion dollar businesses operating across multiple regions. What unifies them is that the problems we are called in for are not problems of building from scratch; they are problems of changing something that has been working for a long time, often in the presence of organisational habits that resist the change.

A few patterns recur.

Growth has slowed and the reason is not obvious. An established technology company arriving at Salamander has usually already noticed the slowdown. What it has not usually done is assessed it accurately. The instinct is often to attribute the slowdown to a sales execution problem when the underlying issue is a product mix that has aged, a pricing structure that no longer reflects the value, a customer base that is consolidating, or an ICP that has drifted. Untangling the actual driver from the visible symptom is one of the most common opening pieces of work for us in this archetype.

Margin is under pressure and the simple cost answers have already been tried. The companies that come to us at this stage have usually already taken at least one round of cost reduction, and the easy reductions are already gone. The work is rarely “find more cost to cut.” It is usually some combination of repricing, repositioning the service mix, restructuring the cost base to support a different operating model, or improving unit economics on the revenue side rather than only on the cost side. Where Finance leads on the assessment, Operations and Go-To-Market usually carry the change.

The operating model is being challenged by a market shift. A common pattern in B2B technology services is the move from project-based delivery to managed services, or from on-premise software to recurring cloud-based revenue. These are not technology decisions; they are commercial and operational ones, and they reshape almost everything: how revenue is recognised, how customers are sold to, how delivery teams are structured, how margins behave, how investors and boards read the business. Salamander engagements at this stage frequently involve helping a leadership team navigate a transition that has already been decided in principle but not yet executed.

The leadership team is established but uneven. Unlike scale-ups, the leadership team in an established company has usually been in place for a while. The challenge is not building it; it is reading it. Most established-company leadership teams have one or two function heads who are right for the next phase of the business and one or two who are right for the phase the business is leaving. Salamander Associates working alongside such teams have to navigate this honestly — the integrity value applies — without overstepping. The recommendation about people belongs to the CEO; the assessment that informs it can be Salamander’s contribution.

The political environment is denser. An established company has investors who have been in the cap table for years, board members who have favoured positions, customers with long-standing relationships, and employees who remember earlier decisions. Decisions at this stage do not just move the business; they reshuffle the political landscape inside and around it. Reading the politics — not navigating them cynically, but understanding them — is part of how we deliver good work without it stalling on contact with the organisation.

How to read an established company in the early weeks. The signals worth attending to are different from the scale-up signals. Look at the rate of change in the leading indicators (slowing or accelerating, and where), at the distribution of growth across product lines or geographies (concentration risk), at margin trends by customer cohort (the aged cohort often masks the new), and at the cadence with which strategic decisions actually get made (a leadership team that takes six months to decide something has a different problem than one that decides too fast).

For you as a Senior Associate, the practical implication is to expect the work to move more slowly than a scale-up but to carry larger consequences when it lands. Decisions in established companies affect more people, more revenue, more capital, and more political relationships. The bar for the quality of the assessment is correspondingly higher, and the discipline of moving from decisions to outcomes (Chapter 13) is correspondingly harder to maintain.

The third archetype is unlike the other two in an important respect: Salamander’s relationship with private capital is not always a client relationship in the direct sense. Sometimes it is. More often it is a relationship that produces engagements with the portfolio companies a sponsor backs. The shape varies between venture capital and private equity, and within private equity it varies further. That variation matters for how the work shows up.

How the relationships actually work.

With venture capital firms, the typical relationship is introduction-based. A VC firm that knows the kind of work Salamander does will introduce us to a portfolio company that needs the help — a CFO-as-a-Service for a Series B that has lost its head of finance, a CRO-as-a-Service for a portfolio company that has stalled in its sales motion — and the engagement, when it happens, is between Salamander and the portfolio company. The VC firm is the connector, not the commissioner. They are not paying us, and they are not the client. Their interest is in seeing the portfolio company helped.

With private equity firms, the relationship is more varied. Some PE relationships look like the VC pattern: the firm makes introductions to portfolio companies, and engagements are then between Salamander and the portfolio company directly. Other PE relationships are deeper. In those, Salamander Associates act as operating partners for the PE firm — embedded in the deal team or the value creation function, working across portfolio companies on the firm’s behalf — without the carry that accompanies a permanent operating partner role. That structural difference matters for how the work runs, and it is worth understanding which shape you are inside before the engagement gets going.

With boards, the relationship is usually more direct. A board can bring Salamander in to support an executive transition, a strategic review, a transaction process, or a piece of independent assessment work the executive team has been asked to commission. In those cases the board (or a specific board committee) is closer to being the direct client, with the executive team running the engagement day-to-day.

A strategic note worth holding alongside all of this. A trusted relationship with a private capital fund — VC or PE — is one of the most valuable forms of network Salamander has. A sponsor that has seen Salamander work well on one of their portfolio companies will introduce us to others. Over time, a relationship that began with a single introduction becomes a steady source of deal flow. Building those relationships well — being useful, being discreet, being the kind of operator a sponsor wants their portcos to meet — is part of the long-game business development of the company.

Three engagement types recur across these relationships.

Value creation. A PE-backed B2B technology company has a thesis about how value will be created over the holding period — margin expansion, recurring-revenue migration, geographic expansion, product portfolio rationalisation, or some combination. The first ninety days after deal close is usually when that thesis gets translated from an investment paper into an operational plan. This is a natural fit for the Salamander model: fractional CFO-as-a-Service, COO-as-a-Service, or equivalent seats can carry one or more of the workstreams directly, and where Associates act as operating partners for a PE firm, value creation work is the shape the engagement would most likely take. Salamander is positioned for this work and actively building relationships with sponsors where it applies.

Transformation oversight. Occasionally a sponsor or board asks Salamander to provide independent eyes on a transformation programme already underway inside a portfolio company. The brief is usually some version of: is this programme on track, what risks are not being surfaced, and what would we do differently if we were running it? This work intersects with the portfolio company’s management team, who are responsible for the programme. The posture is independent but not adversarial — assessing honestly while staying respectful of the team carrying the work.

Transaction readiness. As a portfolio company approaches a transaction — an exit, a recapitalisation, a strategic sale, or a new funding round — the demand for tidy financials, clean operating data, and a credible forward story increases sharply. Salamander engagements support that preparation. CFO-as-a-Service work in the run-up to an exit, for example, frequently includes upgrading financial reporting to the standard a buyer or institutional investor will expect, building the data room, supporting the management presentation, and helping the company tell the forward story in a way that holds up under scrutiny.

A few things to read carefully when private capital is in the picture.

Who the actual client is. Where a sponsor has introduced Salamander to a portfolio company, the client is the portfolio company, and the sponsor’s role is helpful but not contractual. Where Salamander is acting as an operating partner for a PE firm, the client is the sponsor, and the portfolio company is the audience for the work. Reporting lines and priorities differ between the two, and assuming one shape when the other is in play is a common early-engagement mistake.

The time horizons. Engagements with private capital in the picture tend to compress around milestones the operating team may not control — the next board meeting, the fundraise that depends on a clean Q3, the exit window that closes if a key metric does not improve in the next two quarters. The pace of the work reflects those milestones rather than the natural operating rhythm of the portfolio company.

The standard of evidence. Sponsors are paid to be sceptical, and their internal investment committees are paid to be more sceptical still. Recommendations that work in a portfolio company context need to also work in an investment committee context. That sometimes means producing two versions of the same answer — the operational one and the sponsor-facing one — and being clear about the audience for each.

The political topology. A portfolio company’s CEO is not always the originator of the engagement, and is occasionally not enthusiastic about it — particularly when the introduction came from the sponsor rather than from the CEO themselves. Expect a degree of suspicion from the portfolio company in those situations. A senior outside party introduced by their investor can read, from the CEO’s seat, as the investor checking up on them, as a prelude to a more uncomfortable conversation, or as a precursor to a leadership change. The instinct to keep us at arm’s length until the intent is clearer is rational, and Salamander Associates should walk in assuming it is present rather than be surprised by it. The Salamander posture in those cases is the one set out in earlier chapters: clear about why we are there, useful in the work itself, disciplined about not undermining the operating team. Naming the dynamic openly in the first conversation — that we have been introduced by the sponsor, that we are not there to report on the team, that our job is to help the company move forward — usually accelerates the trust-building more than trying to work around the suspicion. Trust gets built by the work, not by the introduction.

For you as a Senior Associate, the practical implication is to clarify early — and often — who the actual client is, what role private capital is playing in the engagement, and which standard the work is being judged against. The substance of the analysis is rarely the hard part. The framing of the work, and the relationships around it, usually are.

Salamander Associates need to be fluent in the business models our clients actually run, because the same presenting problem — growth has slowed, margin is under pressure, the operating model is creaking — has different causes and different remedies in different models. A “growth problem” inside a pure SaaS business is not the same problem as a “growth problem” inside a services-led software business, and the recommendation that helps one will sometimes hurt the other.

This chapter sets out the main models we see across Salamander’s client base, what changes about the assessment when the model changes, and how to read what kind of business you are actually in.

The main models.

Pure SaaS. The business sells software on a recurring subscription, with relatively light implementation and an emphasis on self-service. Revenue is highly recurring; gross margins are typically high (often in the 70 to 85 per cent range); the unit economics turn on CAC payback, gross retention, and net revenue retention. Growth is principally a function of new customer acquisition plus net expansion of the existing base. The dominant operating disciplines are Go-To-Market (the engine of growth) and product (the engine of retention and expansion).

Services-led software. The business sells software, but the implementation and configuration are heavy enough that a significant portion of revenue and people sit in services. Gross margins are mixed (software in the seventies, services in the twenties to forties), and the business has to manage the mix carefully. Growth depends on both new licence or subscription sales and the services capacity to deliver them. The dominant operating disciplines are Go-To-Market, Operations (services delivery), and Finance (mix management).

Managed services. The business operates a service on behalf of customers — running an outsourced function, hosting an environment, managing a platform — usually with software in the loop. Revenue is recurring and contractual but is closer to operating revenue than to subscription revenue. Margins depend heavily on operational efficiency. Growth depends on landing new customers and expanding the scope of service for existing ones, with attention to retention because switching costs are usually high. The dominant operating disciplines are Operations (cost-to-serve, service delivery, capacity), Finance (margin and cash discipline), and Go-To-Market (account management and expansion).

Hybrid models. Most B2B technology companies are not purely one of the above. A scale-up may run a SaaS product with growing services revenue alongside. An established services business may be migrating toward a managed services or recurring software model. A platform business may have multi-sided economics that share characteristics of more than one model. Hybrid is the most common state, and reading the actual mix is part of the early work on an engagement.

What changes when the model changes.

The metric set that matters changes. A SaaS-style assessment built around NRR and CAC payback may be the wrong lens for a managed services business, where the relevant numbers are cost-to-serve, contracted backlog, gross margin by customer cohort, and renewal rate. Importing the wrong metric set is one of the most common assessment errors when moving between client types.

The growth lever changes. In SaaS, growth is usually about the sales engine and the product. In services-led software, growth is also constrained by services capacity — you cannot sign more licences than you can implement. In managed services, growth is partly about landing new customers and partly about expanding scope inside existing ones; the second motion is usually higher-margin and easier to forecast.

The margin profile changes. Pure SaaS businesses have stable, high gross margins that flex slowly. Services-led businesses can see margins move sharply with mix changes. Managed services businesses can see margins move with operational efficiency, contract structure, or customer concentration. A margin programme that works in one model is rarely the same programme that works in another.

The transition risk changes. Many of our most valuable engagements sit inside a transition between models — most commonly, the move from a project-based services business to a managed services or annuity model. These transitions reshape revenue recognition, the sales motion, delivery structure, the margin profile, and how the business reads to investors. They are also where execution-led partnership pays for itself, because the risks are operational, not analytical.

Reading the model in the early weeks.

A useful assessment is to look at three lenses. The revenue ledger — what portion of revenue is genuinely recurring, what portion is project-based, what portion is one-off licence. The cost base — where the heavy people sit (engineering, services, operations, sales) and how cost-to-serve actually behaves with customer count. The customer base — concentration, contract length, renewal rate, expansion rate. Triangulating across the three usually tells you which model the business is in and where it may be drifting.

Drift deserves particular attention. A business that started life as pure SaaS but has accumulated heavy professional services to win enterprise customers is in a different model than its management may still describe it as. A managed services business that has been quietly losing its annuity discipline through one-off contract concessions may be drifting toward a project-based model without anyone naming it. Being explicit about what model the business is actually in — versus the model the leadership team thinks they are in — is sometimes the single most useful early contribution Salamander makes.

For you as a Senior Associate, the practical implication is to do this work consciously on every engagement. Identify the model. Identify the variant. Identify any drift. Identify which metric set, which growth levers, and which margin profiles apply. The rest of the work usually flows more cleanly once that foundation is in place.

A B2B technology company has dozens of metrics it could track. Senior leaders rarely look at all of them; effective ones look at a small number of the right ones for their stage and model. This chapter sets out the economics that consistently matter in our work, and the metrics each of Salamander’s three disciplines turns on.

A handful of metrics carry most of the assessment weight in B2B technology.

Recurring revenue mix. The proportion of revenue that is genuinely recurring — annual or multi-year, with renewal as the default — versus revenue that is project-based, one-off, or services. Recurring revenue is more predictable, more valuable in a transaction, and more durable through a downturn. The mix is also one of the strongest indicators of where the business model actually sits, as Chapter 20 set out. The trajectory of the mix matters as much as the current level: a business moving from twenty per cent recurring to forty per cent is in a different conversation than one moving in the opposite direction.

Gross margin. The bedrock of unit economics. Gross margin tells you how much of every revenue dollar is available to fund Go-To-Market, R&D, G&A, and ultimately profit. Pure software businesses run high; services and managed services lower. What matters most is the trajectory and the variance across customer cohorts and product lines — a company with stable gross margin in aggregate but worsening margin in its newest cohorts has a problem the headline figure is hiding.

Retention. The single most important metric set in recurring-revenue businesses. Gross retention measures whether customers stay; net revenue retention measures whether the dollars retained and expanded exceed the dollars churned. A SaaS business with net revenue retention above 110 per cent sits in a fundamentally different category to one below 100, regardless of how its growth rate compares. Retention is also a leading indicator: customers who will not renew next quarter usually signal it in the quarter before, in their usage, in their support tickets, in QBR engagement, or in their answers to the right open question.

Profitability. EBIT, operating margin, and the Rule of 40 (which at Salamander we frame as the sum of growth rate and operating margin) are the frames we lean on at the scale-up and established levels. We deliberately avoid EBITDA. Depreciation and amortisation are real costs — particularly in businesses that capitalise significant R&D or carry material capex against their operating model — and EBITDA can flatter performance in a way that does not survive proper scrutiny. The bar on profitability is set partly by stage — fast-growing scale-ups can be loss-making and still be healthy — and partly by mix, since different models support different margin floors. The interesting question is rarely “is the business profitable” but “how profitable would it be at the next scale, with the structural improvements we believe are possible.”

Cash. Cash on hand, free cash flow, runway, and the burn rate that connects them. A scale-up that has eighteen months of runway and a clean path to break-even is in a very different position from one that has eight months and is still investing for growth. Cash is also the metric most strongly tied to the timing of major decisions — boards and sponsors plan around it, and many of our engagements are calibrated to a cash milestone the company is moving toward.

The metrics each discipline turns on.

Finance turns on revenue quality and the cash story. Recurring revenue mix, revenue recognition discipline, gross margin and its drivers, the predictability of the forecast, the integrity of the close, runway and free cash flow. The Finance work Salamander does — CFO-as-a-Service, investor-grade reporting, margin and cash control — is built around these.

Go-To-Market turns on growth efficiency and retention. New ARR or new revenue at acceptable CAC payback, pipeline coverage and conversion, sales productivity per salesperson, channel productivity, gross and net retention, expansion rate. The GTM work Salamander does — sales motion, channel performance, marketing narrative, customer success effectiveness — is built around these.

Operations turns on delivery economics and operating cadence. Cost-to-serve, delivery margin, utilisation, the predictability of the delivery cycle, and the cadence with which the leadership team reviews operating performance. The Operations work Salamander does — managed services build, operating cadence installation, scalable execution programmes — is built around these.

Reading metrics in combination.

No single metric tells you whether a business is healthy. A business growing at forty per cent with NRR below ninety-five is hiding a retention problem inside its acquisition strength. A business with high gross margin but a deteriorating mix is hiding a future margin problem. A business with strong cash but slowing growth may be conserving optionality, or it may be quietly underinvesting. The assessment work is to read the metrics in combination and ask the questions the combination raises, rather than to score each one against an abstract benchmark.

A note on the metrics that get attention versus the metrics that matter. Boards and leadership teams under pressure often focus on the metrics that move easily — the top-line growth number, the headcount, the most recent quarter’s pipeline — at the expense of the metrics that move slowly but matter more, like NRR, cohort gross margin, and rolling forecast accuracy. Part of Salamander’s contribution is to keep the slower metrics in the room when the faster ones are dominating the conversation.

For you as a Senior Associate, the practical implication is to be able to construct a credible scorecard for any client within the first two weeks of an engagement, identify which of the metrics that matter are healthy, which are misleading on first read, and which are flashing warnings. Over time, that scorecard becomes a tool you use without consciously building it. In the first few engagements at Salamander, building it consciously is the muscle worth practising.

The previous chapters of Part 3 have covered our three client archetypes, the main business models, and the economics that matter. This final chapter looks at B2B technology companies through a different lens: the recurring pressure points and failure modes that characterise each stage of their growth. Most engagements Salamander takes on involve a company sitting near, or already inside, one of these failure modes.

The lens is useful because the failure mode at one stage is usually not random — it is the consequence of an unaddressed pressure point from the previous stage. Reading where a company sits on this map is one of the more useful early assessment moves an Associate can make.

Around Series A.

The pressure point is the transition from a business that runs on the founders to a business that can begin to run beyond them. The recurring failure modes are predictable. The sales motion that worked when the founders sold every deal stops working when new sellers cannot replicate it. The finance function, which has been bookkeeping plus instinct, is no longer adequate for a board’s needs. Leadership team meetings consist of the founders briefing colleagues rather than the leadership team actually deciding. Operating cadence is whatever the founders say it is, week to week.

The failure modes Salamander is most often called in to address at this stage: an ICP that has never been sharpened, a sales process that does not reliably convert, a forecast that does not survive contact with a board, and a leadership team that is still effectively a founder layer with reports beneath. A fractional CFO-as-a-Service or fractional CRO-as-a-Service can substantially shift the trajectory at this stage, particularly while permanent hires are being recruited.

Around Series B to Series C.

The pressure point is the move from being able to grow at all to being able to grow predictably. The company has usually scaled the founder-led motion as far as it will go and is now trying to build the infrastructure of a larger company. Sales is being built into a team rather than a small group of operators; marketing is being built into a function rather than a founder activity; the finance team is being upgraded; international expansion is being considered or attempted; channel decisions are being made.

The recurring failure modes here are the failures of premature scaling. Hiring ahead of the operating model that can absorb the hires. Opening a new geography before the home motion is fully repeatable. Building a channel sales motion before the direct sales motion is reliable enough for partners to amplify. Building a marketing engine that produces volume the sales motion cannot convert. Most companies in this stage carry at least one of these mistakes; many carry several. The work is often to identify which of them is currently the binding constraint on growth and to address that one before the others.

Around the late scale-up.

The pressure point is the transition from growing fast to growing durably. The company has reached a size where investors, customers, and increasingly the board itself are asking for predictability — and the company has not yet built it. Forecasting accuracy is the visible symptom; the underlying issues are usually deeper. The operating cadence has not caught up with the scale; the leadership team has gaps that were tolerable at a smaller size but are not now; the finance function is producing numbers that are good enough to file but not good enough to run the business on.

Recurring failure modes at this stage include unit economics that have quietly worsened over the last six quarters while growth masked them; a leadership team with one or two members the next phase will not survive with; a recurring revenue mix that has stalled or reversed; a cost base that has been added in good times and is now hard to unwind without losing capability.

Around the established business.

The pressure point is the transition from growing to defending and then to reaccelerating. Growth has slowed for reasons that are no longer obvious — sometimes external, sometimes structural, sometimes the cumulative weight of decisions taken three or four years earlier. Margin is under pressure because the easy efficiencies have been taken and the business model itself may need to change. The leadership team has been in place long enough to have favoured answers, and the political environment around major decisions is denser.

Recurring failure modes here include attributing a slowdown to the wrong cause (sales execution when the real issue is product-market drift; cost when the real issue is revenue mix); attempting a transformation without an operating model change to support it; cutting cost in places that will be needed when the company reaccelerates; tolerating leadership team members past the point at which the next phase requires different people.

A pattern across the stages.

The failure mode at each stage usually has its roots in something that was left undone at the previous one. A company that did not sharpen its ICP at Series A will see the consequences as a converting problem at Series B. A company that scaled headcount past its operating model at Series B will see the consequences as margin pressure at the late scale-up. A company that did not build a forecasting discipline at the late scale-up will see the consequences as a credibility problem with its board at the established stage. Reading backwards from a current pressure point to its origin is often the most useful assessment move an Associate can make in the first weeks of an engagement.

For you as a Senior Associate, the practical implication is to locate every engagement on this map before going deep on the work. Which stage is the company actually in? Which pressure points are the dominant ones? Which failure modes are visible, and which are still latent? Identifying this early often saves three or four weeks of working on the wrong question.

Part 4 covers the first of Salamander’s three disciplines. The Finance work we do sits across four broad areas: CFO-as-a-Service, financial discipline, margin and cash control, and investor-grade reporting. The chapters that follow take each in turn (Chapters 24 through 26) and close with the engagement shapes that recur in our Finance practice (Chapter 27). This opening chapter sets out what falls inside the discipline, where it connects to the other two, and the standard we hold the work to.

The clearest way to describe the scope is to start with the role most of the work runs through: the fractional CFO.

CFO-as-a-Service is the most established offering in Salamander’s Finance practice. A fractional CFO-as-a-Service brings a Senior Associate into a client’s leadership team as the operating CFO for the share of the week the engagement requires — running the finance function, owning the numbers, sitting in the operating cadence, presenting to the board. The other three areas — financial discipline, margin and cash control, investor-grade reporting — are the substance of what a CFO-as-a-Service does inside that seat. They are also engagements in their own right, where a client is not yet at a stage that justifies a full fractional CFO but does need one or more of the pieces.

Financial discipline. The operating cadence, controls, and rhythms that produce reliable numbers. A B2B technology business that cannot trust its own numbers at the close cannot run on those numbers in operating reviews, and cannot put them in front of a board with confidence. Building or repairing this discipline is the foundational Finance work we do, and it is the subject of the next chapter.

Margin and cash control. The levers that move profitability and the disciplines that move cash. Pricing, mix, cost-to-serve, capital allocation, working capital, runway management. In every model — pure SaaS, services-led, managed services, hybrid — the specific levers differ, but the discipline of running them deliberately rather than reactively is the same. Chapter 25 sets out how we do it.

Investor-grade reporting. The numbers, the narrative, and the management materials that a sophisticated audience — a board, a private equity sponsor, a strategic acquirer, an institutional investor — will rely on without hand-holding. The bar is higher than internal reporting. The work is to build the cadence, the controls, and the integrity of the data feeding the reporting before the reporting itself. Chapter 26 covers it.

A note on the metrics we use and the ones we do not. Salamander uses EBIT and operating margin, not EBITDA, when measuring profitability. The reason is the one set out in Chapter 21 — depreciation and amortisation are real costs in any business with R&D capitalisation or material capex, and EBITDA can flatter performance unhelpfully. When you see EBITDA in a client’s internal materials, the work usually includes reframing the conversation around EBIT and the underlying drivers of the gap between the two.

Where Finance connects to the other two disciplines. Finance is rarely a standalone discipline at Salamander; the most valuable engagements are the ones in which the Finance work is in conversation with the Go-To-Market work and the Operations work. A margin programme that improves the P&L but breaks the sales motion has not done its job. A revenue-recognition redesign that suits the auditor but does not survive the operating cadence has not done its job either. The chapters that follow this one treat Finance as a discipline in its own right, but the engagements they describe almost always reach into the other two.

The standard we hold the work to. Finance work at Salamander is investor-grade by default. That phrase carries some weight inside the company. It means numbers that survive the scrutiny of a sceptical board member, a sponsor’s value creation lead, a buyer’s diligence team, or a strategic acquirer’s CFO. It means a forecast that an Associate would put their name to in front of any of those audiences. It means a close discipline tight enough that the numbers do not move materially in the days after the close is signed off. We hold ourselves to this standard whether the engagement is named CFO-as-a-Service or one of the narrower scopes, because it is the standard the work has to clear to be useful.

For you as a Senior Associate working in Finance engagements, the practical implication is to read the discipline scope of any engagement against the four areas above, and to be explicit early about which the engagement is signed up to do and which it is deliberately not. Most Finance engagements involve more than one area, but rarely all four. Knowing the difference is part of scoping the work well.

Most B2B technology companies Salamander works with do not arrive without finance discipline; they arrive with finance discipline that has not kept pace with the size of the business. The numbers are produced. The board pack is built. The audit gets signed. What is missing is the operating cadence and the control environment that lets leadership actually run the company on the numbers — making decisions on them in real time, putting them in front of an investor without weeks of preparation, surviving the change that scale brings.

This chapter is about how we build that discipline.

Five things sit inside the discipline.

A close that closes on time. The monthly close should land on a predictable day each month, with material accuracy, with reconciliations done, and with the numbers stable enough that they do not move materially in the days that follow. Many of the companies we work with arrive with a close that drifts — sometimes three days late, sometimes a week, sometimes a month at year-end. The drift is usually a symptom of unclear ownership, manual processes that have outgrown the tooling, or a finance team that is firefighting other priorities by the time the close window arrives. Stabilising the close is often the first piece of finance work an engagement does, because everything downstream depends on it.

A rolling forecast that is actually used. A static annual budget that is reviewed quarterly is not a forecast; it is a budget. A rolling forecast — usually thirteen weeks for cash and twelve to eighteen months for the P&L, updated monthly — is what lets a leadership team see what is coming and adjust. The forecast has to be owned by Finance but informed by the operating leaders whose numbers feed into it. Building this rhythm is one of the most leverage-rich pieces of work in early Finance engagements: the same forecast that nobody is using becomes a tool the leadership team brings out in every operating review.

Controls that match the size of the business. Controls that fit a Series A scale-up do not fit a company three times its size. We do not import a control framework from a larger company and force-fit it. Instead, we look at where the actual risks live in the business — revenue recognition complexity, cash handling, expense approvals, payroll, system access — and build controls proportionate to those risks. Over-controlled environments slow the business unnecessarily; under-controlled environments break visibly at the next audit or due diligence. The judgement of where the line sits is part of the work.

KPIs and operating metrics that drive behaviour. The metric set that matters varies by model, as Chapter 21 set out. Building the discipline involves defining the right set, instrumenting the data sources so they can be produced consistently, and embedding the metrics in the operating rhythm so they actually shape decisions rather than appearing as appendices in the board pack. The instrumentation is rarely glamorous. It is also rarely optional.

An operating cadence that makes the numbers useful. None of the above matters if leadership is not running operating reviews against the numbers in a disciplined cadence. The monthly business review, the quarterly forecast update, the weekly cash review where appropriate, the board cycle that pulls it all together — these are the rituals that turn finance discipline from a documentation exercise into operating leverage. We often install or refresh this cadence as part of a Finance engagement.

How the work proceeds.

The first weeks of a Finance discipline engagement are usually about reading what is in place rather than installing what is not. An assessment of the close, the forecast, the controls, the KPIs, and the cadence — where is each one healthy, where is each one fragile, and where is each one missing entirely. The Senior Associate produces a short, clear read on this — usually a memo, not a deck — and uses it to scope the work that follows.

What follows is iterative, not waterfall. We pick the area where the discipline is most fragile, install or repair it, and move to the next. We do not build a comprehensive new finance operating model in one go, because the team running the function still has to close the books while the work is happening. The discipline lands one piece at a time, and the leadership team sees the difference in the operating reviews as it lands.

A note on team capability. Financial discipline does not live in the spreadsheets or the systems. It lives in the people running the function. Part of the Finance work Salamander does is reading the existing finance team — who has the capability to run the discipline once installed, who can be developed into the role, and where a permanent hire is the right answer. The recommendation about people belongs to the CEO, as Chapter 18 set out; the assessment that informs it is part of our contribution.

For you as a Senior Associate working in Finance engagements, the practical implication is to think of financial discipline as a system rather than a list. The five pieces above interlock — a close that closes on time enables a forecast that is actually used; controls that match the business enable a cadence that runs cleanly; KPIs that drive behaviour are the connective tissue across the others. Building any of them in isolation rarely sticks. Building them in coordination is how the discipline actually takes root.

Margin and cash are the two financial outcomes the rest of Salamander’s Finance work feeds into. A company that has installed financial discipline (Chapter 24) and is preparing investor-grade reporting (Chapter 26) is producing the numbers that let leadership and the board see what is happening. This chapter is about acting on what the numbers show — the levers that actually move margin and cash, the order in which they tend to be most effective, and the discipline that makes the improvements stick rather than reverse the moment attention shifts.

A note before the levers. Most companies arriving at Salamander with a margin problem have already done the obvious things. Headcount has been reviewed. Discretionary spend has been trimmed. A budget exercise has produced some efficiency. The work we are usually called in to do is the work that comes after that — where the simple answers have run out and the underlying drivers have to be addressed.

The margin levers, in roughly the order they tend to be most effective.

Pricing. The single highest-leverage margin lever in most B2B technology businesses is pricing, and it is the one most companies under-use. List prices that have not moved in years. Discounting that has crept up across the customer base. Pricing structures that no longer reflect the value the product delivers. Packaging that has fragmented and become hard to sell against. Sharpening pricing — sometimes through a price increase, sometimes through repackaging, sometimes through tightening discounting discipline — frequently produces more margin improvement than a year of cost work, with less organisational pain. We push to look at pricing first.

Revenue mix. The blend of recurring versus one-off revenue, of higher-margin versus lower-margin products, of customer segments with different unit economics. Margin programmes that shift the mix toward the higher-margin pieces compound over time, where margin programmes that only attack cost are one-shot. Mix shifts often run through Go-To-Market (which customers to pursue, which products to lead with) and through Operations (how delivery cost is structured), which is why margin work rarely sits inside Finance alone.

Cost-to-serve. The fully-loaded cost of serving each customer, segment, or product line. Most companies do not know their cost-to-serve at the level of detail required to act on it; getting to that view is often the most leverage-rich assessment move in a margin engagement. Once visible, the work usually involves repricing the customers where cost-to-serve is too high, redesigning the delivery model where the cost structure is wrong, or rationalising the bottom decile of customers where neither of the first two will work.

Fixed cost structure. Headcount, facilities, systems, vendor spend, professional fees. The simpler answers have usually been taken; the harder ones involve reshaping the structure rather than trimming around the edges — moving from one delivery model to another, consolidating duplicated functions, replatforming systems that are absorbing cost the business does not need to carry. These changes are slower, more disruptive, and usually higher-impact than incremental cost trimming.

Variable cost discipline. Discounting, give-aways, services scope creep, soft commitments to customers that turn into hard delivery cost. These are not headline budget items; they are the leaks. Surfacing them and tightening the discipline that lets them happen is part of every serious margin programme.

The cash levers.

Working capital. Receivables, payables, inventory where relevant, prepayments. Disciplined collections, sensible payment terms, and proper alignment of cash collection to revenue recognition can free up cash that is sitting in the working capital cycle without changing the underlying business. Most companies have some slack here.

Capital allocation. What the company is spending capital on, and whether the returns justify the spend. Software licences nobody is using. R&D projects that have not been killed. Headcount investments in functions the operating model has outgrown. Cash control includes deciding what to stop, not only what to slow.

Runway management. For scale-ups in particular, runway is the single biggest constraint on what the business can do. Engagements that improve cash burn meaningfully extend optionality and reduce the dependency on fundraising at unfavourable moments. The discipline is to make runway a managed variable in the operating cadence rather than a number that surfaces at the next board meeting.

How we make it stick.

Margin and cash improvements that come from a one-off programme tend to reverse within a year. The improvements that last are the ones that get embedded in the operating cadence — pricing reviewed quarterly, cost-to-serve reported in operating reviews, runway visible at every monthly close, discounting discipline owned by a named individual in the GTM organisation. Part of the Finance work Salamander does is making the cadence carry the discipline forward after the engagement steps back.

For you as a Senior Associate working a margin or cash engagement, the practical implication is to resist the instinct to start with cost cuts. Look at pricing first. Look at mix second. Look at cost-to-serve before fixed cost. Build the cadence to hold the improvements after they are made. And remember that margin work which is technically right but politically incoherent — price increases the GTM team did not buy into, cost cuts in functions that were producing growth — usually unwinds itself within two quarters.

Investor-grade reporting is the third substantive area in Salamander’s Finance practice. It is the area where the financial discipline of Chapter 24 and the margin and cash work of Chapter 25 become visible to the audiences that matter most to a B2B technology company: its board, its investors, its sponsors, and, when the time comes, its acquirers. This chapter sets out what “investor-grade” means in practice, what the audiences actually need, and how we produce it without falling into the trap of overbuilding.

The audiences differ, but they share a few characteristics. They are sophisticated. They have seen many companies. They have limited time to read a pack. They are sceptical by default, and they reward credibility over completeness. They are looking for a small number of signals that tell them whether the business is healthy and whether the leadership team is in command of it.

What each audience actually needs.

Boards need a clear read on the health of the business, the things to worry about, the things that have changed since the last meeting, and the decisions the board itself needs to make. Most board packs over-deliver on the first item, under-deliver on the second and third, and treat the fourth as an afterthought. The Salamander discipline is to invert that distribution.

Investors need the same information at lower frequency, framed against the thesis they backed. A scale-up’s quarterly investor update should answer four questions: how is performance against the plan, what is changing in the operating environment, what are the priorities for the next quarter, and what would the company like from its investors. The narrative is short, the numbers are the right numbers, and the asks are explicit.

Private equity sponsors need a more granular view, calibrated to the value creation plan and to the milestones the sponsor is tracking against. Where Salamander is acting as an operating partner for a PE firm, the reporting is built for the firm’s internal use as much as for the portfolio company’s. Where Salamander is engaged through a CFO-as-a-Service inside a sponsor-backed company, the sponsor sits alongside the board as a primary audience, and the reporting reflects that.

Strategic and financial acquirers need the numbers, the narrative, and the data room. Investor-grade reporting in the run-up to a transaction is reporting that survives buyer-side diligence intact — not because it has been embellished, but because the integrity of the close, the cleanliness of the data, and the coherence of the forward story have all been built to that standard for the months and quarters leading up to the process.

What the reporting actually contains.

A short narrative section that tells the reader what has happened, what has changed, and what to look at. The narrative is what the executive reader actually reads; the numbers are what they refer to. Most companies skip the narrative or treat it as a foreword. We do not.

The headline numbers, presented at the level of altitude the audience needs. Revenue, gross margin, EBIT, cash, runway, and the two or three operating metrics that matter most for the model. Variance to plan, to prior period, and to forecast.

A small number of deeper cuts — the metric trees that show why a top-line number moved, the cohort views, the pipeline views, the regional or segment breakdowns relevant to the moment. The discipline is to choose. A pack that includes everything that might be relevant is a pack that hides the things that actually are.

The risks worth surfacing, named honestly. Investor-grade reporting that lists every conceivable risk is doing the same job as one that lists none — it is not helping the reader. The discipline is to name the two or three things that actually matter, with what is being done about them.

Decisions or asks. What the audience is being asked to react to, decide, or support. Many board packs do not contain this section. Adding it changes the way the audience reads the rest of the pack.

The discipline of not overbuilding.

The most common failure mode in investor-grade reporting is over-engineering. A pack that has grown by accretion over many quarters, full of slides that nobody reads, with the actual signal buried somewhere on page forty. Salamander engagements often begin by stripping such a pack back to a fraction of its size and rebuilding around what the audience actually needs. The pack gets shorter and the meetings get better.

The data feeding the reporting matters more than the reporting itself. A beautiful pack built on data that does not reconcile to the close, or that depends on three days of manual rework each cycle, is worse than a plain pack built on data the team trusts. Investor-grade reporting depends on investor-grade discipline upstream — the work of Chapter 24.

For you as a Senior Associate working on investor-grade reporting, the practical implication is to resist the urge to add. Read what already exists, identify what the audience is actually asking, and edit aggressively. The first time you cut a forty-slide board pack to twelve and watch the meeting work better, the discipline becomes natural. Most of the time, the question is not “what else should we include” but “what can we cut without losing what actually matters.”

The chapters in Part 4 so far have set out the substance of Salamander’s Finance discipline. This chapter is about the shapes those engagements take. Every Finance engagement is in some sense bespoke, but they cluster into a small number of recognisable shapes. Knowing which one you are inside helps you scope the work, calibrate the pace, and read what the client actually needs.

The list is not exhaustive. Patterns evolve with client demand, and new shapes appear. The two below are the most recurrent at the time of writing.

Shape 1 — Improve profitability.

The most common Finance engagement shape. A B2B technology company has a margin problem, a cash problem, or some combination of the two, and is looking for a Senior Associate to lead the work to fix it. The engagement typically opens with an assessment that establishes which of the levers from Chapter 25 are the right ones to pull — pricing, mix, cost-to-serve, fixed cost, variable cost discipline, working capital, capital allocation, runway — and then moves into installing the programme of work that pulls them.

This shape is Finance-led but rarely Finance-contained. A pricing change runs through Go-To-Market. A cost-to-serve programme runs through Operations. A working capital improvement runs through both. The Salamander Associate carrying the Finance seat is usually working alongside other Associates carrying the GTM or Operations work in concert. The discipline is to keep the financial outcome — the margin or cash improvement — visible at the centre, while the levers themselves are pulled across more than one function.

Engagements of this shape run anywhere from three months to two years, depending on the scale of the company and the depth of the programme. The pace varies more than most shapes. A profitability engagement inside a scale-up with eight months of runway moves at a different speed from one inside an established company working through a multi-year operating model migration.

Shape 2 — Prepare financials for exit or fundraising.

The other recurrent Finance shape. A B2B technology company is heading toward a transaction — an exit, a recap, a strategic sale, or a new round of investment — and needs the financials, the operating data, and the forward story to land at the standard the audience will expect. The engagement typically combines elements of Chapters 24 and 26 — closing the gaps in the underlying discipline that the data depends on, and building the reporting the transaction audience will consume.

This shape has a clearer endpoint than the profitability one. The transaction milestone sets the timeline; the work compresses around it. A CFO-as-a-Service Associate in this shape of engagement is typically the most visible Salamander presence at the company in the months leading up to the process — running the close to higher standards, building the data room, supporting the management presentation, sitting alongside the CEO through investor or buyer meetings.

The judgement call most often required in this shape is what to fix in time versus what to flag and accept as a known gap. A six-month run-up to an exit is not a six-month period in which the entire finance function can be rebuilt; it is a period in which the work has to be ruthlessly prioritised around what the buyer or investor will actually scrutinise.

Other shapes worth knowing.

The standalone fractional CFO-as-a-Service, holding the seat while a permanent hire is recruited or developed. The finance discipline overhaul, focused on Chapter 24’s work without the margin or transaction angle. The early-stage budget build — typically commissioned after a funding round, when investors expect to see an operating budget the board can run against. The engagement builds the budget from the company’s post-round plan and milestones, sense-checks the assumptions, and puts the result into a form the new investors will accept. These engagements often become the start of a longer CFO-as-a-Service relationship as the company grows into its post-round operating discipline. The finance systems migration — scale-up technology companies often outgrow the finance systems they started on (commonly QuickBooks, Xero, or similar), particularly as they move into multiple countries or currencies. Salamander is building a capability to help them lift and shift to more appropriate systems such as NetSuite. This is an emerging shape rather than an established practice; check the current state of the capability before scoping a migration engagement. The board pack rebuild, where the reporting is the engagement. The runway extension programme, where cash and capital allocation are the focus. The carve-out or integration finance support, where a transaction has happened and the finance function on one side or the other needs reshaping. The post-investment value creation engagement, often initiated through a PE relationship and described in Chapter 19.

A note on how shapes blur in practice.

Many Finance engagements begin as one shape and develop into another. A profitability engagement runs into a structural issue that turns it into an operating model question. A board pack rebuild surfaces a discipline gap that turns it into a finance function upgrade. A CFO-as-a-Service engagement is informed in month four that the company is going to be sold, and a transaction-readiness scope is added on top. The shape tells you where you start. It rarely tells you where you finish, and the willingness to reshape the engagement as the work develops is part of the practice.

For you as a Senior Associate, the practical implication is to name the shape on day one and to keep naming it as the engagement evolves. A clear shared view of the shape of the engagement — between you, the Associate leading it, and the client — is one of the things that prevents scope drift later. When the shape genuinely changes, name that too, and reset the conversation rather than letting the engagement drift inside its original framing.

Part 5 covers the second of Salamander’s three disciplines. Go-To-Market at Salamander spans five elements: sales, channel sales, international growth, marketing, and customer success. The chapters that follow take each in turn (Chapters 30 through 34) and bookend them with an assessment chapter (29) and an engagement shapes chapter (35). This opening chapter sets out what falls inside the discipline, how the five elements fit together, and the standard we hold the work to.

The first thing to say is that the five elements are not five disciplines. They are five parts of a single GTM system. A company that has fixed its sales motion but failed to align marketing to it has not fixed its GTM. A company that has built a strong marketing engine but cannot retain the customers it acquires has not solved a GTM problem. The work Salamander does in GTM is almost always about how the five interact, not about any one in isolation.

The five elements.

Sales. The direct motion through which the company sells to customers. The motion may be human-led (a sales team carrying named accounts), product-led (customers signing up and converting through the product itself), e-commerce-led (online purchase without a sales conversation), or a hybrid of the three. Pipeline, forecasting, sales motion design, conversion economics, segmentation, sales operating cadence — and, where the motion is human-led, salesperson productivity and sales team design. The seat the work typically runs through is CRO-as-a-Service. Chapter 30 sets out the substance.

Channel sales. The motion through partners, resellers, integrators, and other indirect routes to market. Channel design, partner programmes, channel productivity, the balance between direct and channel motions. As Chapter 22 noted, channel works as an amplifier of a working direct motion, not as a substitute for one. Chapter 31 goes deeper.

International growth. The motion through which the company expands into new geographies. Market selection, entry models, the operational lift required in each new market, sequencing of geographic expansion. The work intersects with sales (often a localised version of the direct motion), with channel sales (channel is often the right entry model in markets where direct does not yet scale), and with Operations (delivery in new markets has its own challenges). Chapter 32.

Marketing. Positioning, messaging, demand generation, brand at B2B technology, content, and the handover from marketing to sales. The seat the work typically runs through is CMO-as-a-Service. Chapter 33.

Customer success. The motion that holds, expands, and renews existing customers. Retention, NRR, expansion motion, customer health, and the recurring revenue mix that flows from all three. Chapter 34.

How the five fit together as one system.

A useful way to read a B2B technology company’s GTM is as a single funnel from market awareness through to renewal and expansion, with each of the five elements responsible for a stage of that funnel.

Marketing creates the awareness and the demand that produces the pipeline. Sales converts the pipeline into customers. Channel sales is either an amplifier of that motion or a parallel motion serving segments the direct motion cannot reach efficiently. International growth opens new geographies for the same end-to-end funnel. Customer success retains and expands the customer base once acquired, producing the recurring revenue that compounds.

When the system is working, the elements reinforce each other. When it is not, the failure is often visible at one element but caused by another — pipeline that is poor because marketing is not producing it; sales productivity that is poor because customer success is producing churn faster than sales can replace; channel underperformance that turns out to be a direct-motion problem in disguise.

Assessing across the system, rather than into a single element, is the work that gets the right answer. The next chapter is about how we do that.

Where GTM connects to the other two disciplines.

GTM is the engine of growth, but the engine runs on inputs from the other disciplines. The unit economics of the sales motion — CAC, payback, mix — are Finance’s responsibility to make visible. The delivery model that absorbs new customers without breaking sits in Operations. A GTM engagement that does not stay in contact with Finance and Operations is a GTM engagement that will produce work that does not survive contact with the wider company. The work runs alongside, not in a silo.

The standard we hold the work to.

GTM work at Salamander is held to the same standard the rest of the practice is — investor-grade, decision-ready, and outcome-focused. A pipeline view that does not survive a sceptical board member is not a pipeline view we will sign our name to. A growth plan that cannot be defended against a sponsor’s diligence is not a plan we will recommend. A CRO-as-a-Service or CMO-as-a-Service Associate is operating to that bar, in real time, while the work is being done.

For you as a Senior Associate working in GTM engagements, the practical implication is the same as in Finance: read the discipline scope of any engagement against the five elements above, and be explicit early about which the engagement is signed up to do and which it is deliberately not. Most GTM engagements involve more than one element, but rarely all five. Knowing the difference — and naming it — is part of scoping the work well.

A GTM assessment at Salamander is usually the first piece of work on a GTM engagement, regardless of which of the five elements the engagement is ultimately scoped against. The reason is the one Chapter 28 set out: the visible symptom is usually at one element, but the binding constraint is often somewhere else in the system. Getting the assessment right saves the engagement from working on the wrong question.

This chapter sets out how we run that assessment.

The funnel as the underlying frame.

Read the company’s GTM as a single funnel: market → awareness → demand → pipeline → conversion → customer → renewal → expansion. Each stage has metrics, and each stage has a typical failure mode. The assessment is to walk the funnel end to end and identify where the leak, the constraint, or the structural problem actually sits.

In a human-led sales motion, the funnel runs through marketing, qualification (often through a business development or inbound function), pipeline conversion by salespeople, customer onboarding, retention, and expansion. In a product-led motion, the funnel runs through marketing, sign-up, activation inside the product, conversion to a paid plan, retention, and expansion. In an e-commerce motion, the funnel runs through traffic acquisition, conversion on the site, purchase, retention, and expansion. The stages have different names; the assessment logic is the same.

What we look at in the first two weeks.

The data first. Pipeline coverage — both unweighted (the raw view: are there enough opportunities) and weighted (probability-adjusted: is the number likely to land) — conversion rates by stage, cycle times, win rates, CAC and CAC payback, gross retention, net revenue retention, customer cohorts, expansion economics. The point at this stage is not to score the company against industry benchmarks but to find the discontinuities — the stage where the funnel narrows sharply, the cohort where retention falls off, the segment where conversion stalls.

The motion next. Sit in a pipeline review or a customer success review. Listen to a few sales calls if the motion is human-led. Walk the product onboarding if it is product-led. Open the website and try to buy something if it is e-commerce-led. The data shows the symptoms; the motion shows the cause.

The team after that. Who owns growth at the leadership level — is it a single person, or is the responsibility distributed across functions that are not coordinated? What is the cadence with which growth is reviewed? Does marketing know what sales is forecasting? Does customer success know what marketing is promising? Does the sales team understand how the product is supposed to be sold? Misalignment across these functions is one of the most common GTM problems, and it does not show up in any single metric.

The strategy underneath. What is the ICP, and is it actually the customer the company is winning? What is the value proposition, and does it land with that ICP? What is the pricing structure, and does it reflect the value delivered? GTM problems are sometimes execution problems and sometimes positioning problems, and the assessment has to be able to tell the difference.

The recurring binding constraints.

Across many engagements, a few binding constraints recur, and recognising them quickly accelerates the assessment.

ICP drift. The company is winning customers who are not the ICP, with elongated sales cycles, lower retention, and higher cost-to-serve. The fix is to sharpen the ICP, but the symptom is usually visible at conversion or retention rather than at targeting.

Pipeline starvation. Marketing is not producing enough qualified pipeline for the sales motion to convert. The sales team is blamed for missed numbers, but the constraint sits upstream. This is one of the most common assessment errors in B2B technology.

Retention leak. Customer success is producing more churn than sales can replace at the rate the company is investing in acquisition. Net revenue retention is the canonical metric; the underlying drivers are usually some combination of fit, onboarding, product reliability, and account management.

Sales motion at the wrong altitude. The motion that worked at the previous size of the company has not been redesigned for the current size. Founders are still selling deals the sales team should be running; senior salespeople are still doing prospecting work that an automated or junior qualification motion should be doing. The fix is rarely simple but the assessment usually is.

Channel mismatch. Either the company is leaning on channel to compensate for a direct motion that does not work, or it is failing to use channel to amplify a direct motion that does work. Chapter 31 goes deeper on this.

The output of the assessment.

A Salamander GTM assessment produces a short, written read — usually a memo with one or two supporting exhibits, not a deck — that names the binding constraint, the underlying drivers, the data that supports the assessment, and the work that follows. The point is to converge on a clear answer, not to inventory every possible issue. The longer the assessment memo, the less useful it usually is.

For you as a Senior Associate, the practical implication is to keep the funnel as your backbone and the binding-constraint question at the front of your mind. Walk the funnel. Find the discontinuity. Identify what is upstream of it. Triangulate across data, motion, team, and strategy. Most useful GTM assessments converge in two to three weeks; the cases where they do not are usually cases where the company has more than one binding constraint at once — and even then, naming which one to address first is the most valuable contribution you can make.

Sales is the part of GTM that converts demand into revenue. As Chapter 28 noted, that conversion can run through different motions — human-led, product-led, e-commerce-led, or hybrid — and the design of the motion is the first thing a Salamander Associate has to be clear about before getting into the substance of the work.

This chapter sets out the substance: how we read the sales motion, what we look at, what we build, and what we install.

Sales motion design.

The motion is the architecture of how revenue gets won. In a human-led motion, the questions are about whether the sales process is reliable enough to be run by anyone other than its current top performers. In a product-led motion, the questions are about whether the product itself converts users to customers at acceptable rates, and where human intervention is needed alongside it. In an e-commerce motion, the questions are about whether the site experience moves visitors to purchase with the right basket value, the right repeat rate, and the right unit economics. In a hybrid motion, the questions are about how the elements interact, where customers move between them, and whether the handovers work.

The Salamander work in motion design is to make the motion fit the company’s stage, customer segment, and product. A motion that is over-engineered for the stage (a heavy enterprise sales process at Series B for a deal size that does not support it) or under-engineered (a transactional motion for an enterprise sale that needs more weight behind it) is a motion that will not produce predictable revenue. Calibration is part of the work.

Pipeline discipline.

Pipeline is the leading indicator of revenue, but only if it is run with discipline. In our work, pipeline discipline means three things in combination: pipeline that is honestly stage-gated (an opportunity at Stage 3 actually has Stage 3 evidence behind it); coverage that is read both unweighted and weighted; and a cadence that walks pipeline weekly with the discipline to remove what is not real, not just add what is new. Most pipeline reviews we walk into in early-engagement weeks suffer from one or more of these gaps. Tightening them is foundational to everything else.

Forecasting accuracy.

A forecast is only as useful as it is accurate. Forecasting accuracy at Salamander is measured not by how close the headline number is, but by how stable the forecast is across the cycle — does the same opportunity move between commit, best case, and pipeline three times before close, or does it stay where it was first placed by a salesperson who has read it correctly? Stability across the cycle is the signal. Companies with high forecasting accuracy tend to have high revenue predictability; companies without it tend to discover bad news late.

Revenue predictability.

The destination this work is heading toward. Revenue predictability means a leadership team can tell its board, with confidence, what next quarter and the quarter after look like, and be substantially right. Predictability is the product of pipeline discipline, forecasting accuracy, motion design that is mature for the stage, and a salesperson productivity profile that holds across cycles rather than swinging from quarter to quarter. Building all of these is what a CRO-as-a-Service engagement is, in operational terms, signed up to deliver.

Salesperson productivity and sales team design.

Where the motion is human-led, two questions matter. How productive are the salespeople carrying named accounts — measured by quota attainment, ramp time, productivity at full ramp, and consistency of performance across the team. And how is the team designed — what is the right ratio of new-business salespeople to account managers, what coverage model fits the customer segments, where is the line between an inside-sales motion and a field motion, how is qualification handled before opportunities reach a closing salesperson. Both are recurring areas of Salamander work. Assessing both at once is how we usually start.

Sales operating cadence.

The rituals that hold sales together. The weekly pipeline review. The monthly forecast review. The quarterly business review with the leadership team. The annual planning cycle. Many of the Salamander engagements that work well install or refresh this cadence in the first month, because the operating rhythm is the structure inside which everything else gets done. A sales team without a disciplined cadence cannot run a disciplined pipeline; a disciplined cadence makes the other work possible.

Common patterns.

A few patterns recur in sales engagements at Salamander.

The motion has outgrown its design. The company has scaled past the size at which the original motion worked, but the motion itself has not been rebuilt. New salespeople join and underperform because the process they are running is not designed for the company the business has become.

Forecast accuracy that does not survive scrutiny. The headline forecast lands often enough that nobody questions it, but the stage-gating underneath is loose, and the next time the macro environment shifts the forecast collapses. The work is to tighten the discipline before the environment forces the issue.

Concentration on a small number of top performers. Two or three salespeople carry most of the number; the rest of the team is below quota. The assessment is usually a motion that requires more skill than the average salesperson on the team has — meaning the right answer is not always “hire better people,” but to redesign the motion so a competent salesperson can succeed at it.

For you as a Senior Associate working in a sales engagement, the practical implication is to anchor everything to predictability of revenue. Pipeline that does not produce predictable revenue is not pipeline that matters. A forecast that does not hold across the cycle is not a forecast worth maintaining. The work of the engagement is to move the company from where it is to where its leadership and board can rely on the numbers.

Channel sales — selling through partners, resellers, integrators, marketplaces, and other indirect routes to market — is one of the most powerful levers a B2B technology company has, and one of the most commonly mis-used. This chapter sets out how we think about channel, how we assess channel performance, and how Salamander engagements typically address channel issues.

The first principle, set out in Chapter 22 and worth reinforcing here, is that channel works as an amplifier of a working direct motion, not as a substitute for one. If the company’s own salespeople cannot reliably win deals, partners will not be able to win them either, and the margin paid to partners simply reduces the unit economics of the deals that do happen. The assessment question to ask before any channel work is: does the direct motion work well enough to be amplified?

Once that question is answered honestly, channel work has four main dimensions.

Channel design.

The first question is what channel a company actually needs. A few common shapes recur. Resellers and distributors who carry the product to segments the direct motion cannot reach economically — typically smaller customers, geographies without a direct presence, or industries with established intermediary relationships. System integrators and consulting partners who attach the product to larger transformation programmes and bring it into accounts with their own relationship. Technology and ISV partners whose products extend or complement the company’s, often through a marketplace or co-sell relationship. OEM and white-label arrangements where the product is embedded in another company’s offering. Marketplace channels (cloud marketplaces, app stores, ISV exchanges) where the transaction is digital.

Most B2B technology companies arrive at Salamander with some mix of these. The assessment is rarely “build a channel from scratch”; it is more often “rationalise the channel programme down to the partners that actually produce, and invest in the ones that do.”

Partner programmes.

A channel works because partners find it commercially attractive and operationally easy to sell. The programme that makes this happen has a few components: clear economics for the partner (margin, rebates, MDF, deal registration); enablement that lets the partner sell the product without continuous hand-holding; technical support that makes implementation viable; and a partner-management function on the company side that maintains the relationship.

Most companies under-invest in one or more of these. The most common shortfall we see is enablement — partners who are nominally in the programme but never reach a level of confidence in the product or the pitch to sell it independently. The second most common is the partner-management function — partners signed up at a conference and then never spoken to again. The fix is rarely glamorous: a smaller number of partners, each genuinely enabled and supported, will produce more revenue than a long list of paper partners.

Channel productivity.

Measuring channel performance is its own discipline. Productivity by partner (revenue generated, pipeline contributed, deal size, win rate), productivity by partner segment (resellers versus integrators versus marketplace), and productivity per dollar of programme investment. Most channel programmes are running on instinct rather than on data; getting the data in place is often the first move on a channel engagement.

The hardest piece is attribution — a deal that closes through a partner often had marketing, the direct sales team, and the partner involved at different points. Untangling who is responsible for what is part of getting honest about which channels are producing and which are absorbing investment without returning it.

Direct/channel balance.

Where the direct motion and the channel motion both reach a customer, the rules of engagement matter. Channel conflict — two routes competing for the same deal — destroys partner trust and consumes management attention. Most B2B technology companies have some version of this issue at some scale; the work is to design the segmentation and the rules so the conflict is rare and, when it happens, the resolution is predictable.

Deal registration is the standard mechanism. A partner who identifies and qualifies an opportunity registers the deal with the company; once registered, the partner is protected from direct sales pursuing the same opportunity, and from other partners registering it on top, for an agreed period. In return, the partner commits to working the deal to a defined standard and reports progress against it. Well-run deal registration sits inside the partner portal, is governed by clear and published rules (what qualifies for registration, how long protection lasts, what triggers expiry, who adjudicates disputes), and is connected to the company’s CRM so direct sales can see what is registered before they pursue an account. A company without a working deal registration programme is, in effect, asking its partners to trust that channel conflict will be handled fairly case by case — and partners almost never extend that trust for long. Where Salamander engagements involve channel design or channel repair, getting deal registration right is one of the most consistently high-leverage moves available.

A useful test: ask three partners what they would do if a direct salesperson appeared on one of their deals. If the answer is consistent across the three, the rules work. If the answers diverge, the company has a channel conflict problem regardless of what its policy documents say.

Common failure modes.

A few patterns recur.

Channel built before direct works. The original mistake, addressed in Chapter 22. The work is usually to scale the channel programme back, fix the direct motion, and rebuild the channel on a working foundation.

Long tail of paper partners. Two hundred signed partners; ten of them produce ninety per cent of the channel revenue; the other one hundred and ninety absorb programme cost without returning it. The fix is to rationalise — invest in the ten, formally tier the next layer down, and let the rest fall away through inactivity.

Channel margin given up without channel value received. Partners receive significant margin but contribute little beyond order-taking that the direct motion could have done. The fix is either to redesign the relationship so the partner adds genuine value, or to bring the customer back direct and absorb the partner-management cost saving.

Conflict that has been left unaddressed. Direct and channel are competing on the same accounts, partners are protesting, the company is paying margin twice on some deals and losing partner trust on others. The fix is segmentation and clear rules of engagement, and it is usually overdue by the time we are called in.

For you as a Senior Associate working on a channel engagement, the practical implication is to start with the direct motion. Confirm it works. Then look at the channel programme honestly — productivity by partner, programme economics, the partner management function. The work is more often about doing less with fewer partners better than about building a bigger programme.

International expansion is one of the most consequential decisions a scaling B2B technology company makes, and one of the most commonly mistimed and mis-executed. Salamander engagements on international growth divide roughly into three questions: when to expand, where to go, and how to enter. Each has its own discipline, and getting any one of them wrong is enough to make the whole exercise more expensive than it needed to be.

When to expand.

The instinct of many scale-up CEOs is to expand internationally as soon as the home market starts to slow. That is usually the wrong moment. International expansion succeeds when the company is exporting a motion that already works at scale in its home market, and it fails — slowly and expensively — when the company is exporting a motion that does not yet work at home. The first assessment question on any international engagement is whether the home motion is repeatable and predictable, or whether the company is hoping international markets will produce the growth its home market is not.

A second readiness question is leadership bandwidth. International expansion absorbs disproportionate senior attention in the first eighteen months. If the leadership team is already stretched by the home market, the right answer is rarely to add the cognitive load of a new geography on top.

Where to go.

Market selection in B2B technology is rarely about which geography is largest. The variables that matter are the size of the addressable market for the company’s product specifically, the strength of the ICP fit in that market, the maturity of the buyer category, local competitive density, the cost of operating, the proximity of the home leadership team for time zone and travel, and the presence of any existing customers or channel relationships that can anchor a new presence.

A common pattern in our engagements: the “obvious” market — the largest economy in a region, or the market the founders happen to know best — turns out not to be the right first market. A smaller market with a sharper ICP fit, an existing anchor customer, and lower entry cost often produces better unit economics in the first eighteen months.

For Salamander’s geographies in particular, the answer also depends on whether the company is moving within Asia Pacific, into Europe, into the Middle East, or across all three. Each has its own dynamics. APAC is not a single market; the differences between Singapore, Japan, Australia, India, and China are larger than the differences between most European countries. EMEA shares more linguistic and regulatory commonality but has its own complexity around language localisation, employment law, and currency.

How to enter.

Several entry models exist, and choosing the right one is one of the larger decisions in an international engagement.

Direct presence is the most committed model — incorporate locally, hire a local team, build the operating infrastructure. Most expensive, highest control over the customer relationship and the unit economics. Right where the company has already validated demand and is moving to scale it.

Channel-led entry is the most capital-light model — a local reseller, distributor, or system integrator carries the product to local customers. Low capital outlay, low visibility into customer and unit economics. Often the right way to test demand before committing direct, with the same caveat that channel amplifies a working direct motion better than it substitutes for one.

A hybrid model is increasingly common — a small local presence (often a single senior salesperson or country lead) supported by channel partners who carry the volume, with the local presence handling key accounts and channel management. It works well for many B2B technology companies in their first international moves.

Acquisition or joint venture is occasionally the right answer where the local buying culture makes a wholly foreign company difficult to land. These are larger commitments and outside the scope of most Salamander engagements.

The operational lifts that follow.

The decisions above are the visible part of international expansion. The operational lifts that follow are usually larger than the client expected: a local legal entity with statutory accounting, audit, and tax obligations the home finance team has not previously managed (CFO-as-a-Service work, Part 4); a local employment arrangement, whether through a hired entity, an employer of record, or a contractor model; local pricing calibrated to local willingness-to-pay rather than translated from home pricing; customer success and support running on the local time zone, with the local language where required; and marketing that is localised in tone and channel, not translated. Many international expansions fail more visibly on marketing than on sales.

A specific finance-system implication is worth surfacing here. The multi-country, multi-currency complexity of an international move frequently exposes the limits of the company’s existing finance system — most often QuickBooks, Xero, or similar — and is one of the most common moments at which the finance systems migration covered in Chapter 27 gets triggered. International growth engagements should therefore be read against the finance side from the start: the migration question often surfaces before the company has named it.

Sequencing.

The order in which the lifts happen matters. Successful expansions stand up the bare minimum — entity, first hire or partner, a defensible pricing approach, a way to deliver — and add capability as the revenue comes. Less successful expansions try to build the local company at scale before the revenue justifies it.

Common failure modes.

Going international before the home motion works. The single most common mistake, and the one Salamander engagements are most often called in to correct.

Choosing the wrong first market for reputational reasons rather than economic ones. The CEO wants a US presence because investors will reward it, regardless of whether the unit economics will support it for two years.

Underestimating the operational lift. The country lead is hired before the entity is incorporated; the entity is incorporated before the tax structure is understood; the first customer is signed before the contract template is localised. Months of catch-up work consume the new market’s first-year contribution.

Trying to run the international business from the home country without local seniority. The home leadership team takes the international remit on top of its existing load and discovers it does not have the capacity. A fractional CRO-as-a-Service or country lead — the shape Salamander provides — is often the right answer in the first eighteen months.

For you as a Senior Associate working on an international engagement, the practical implication is to slow the conversation down. International decisions made quickly are often made expensively. The work is to make sure the readiness question has been answered honestly, market selection has been done on real economics rather than reputation, the entry model fits the readiness, and the operational lifts are sequenced in an order the company can execute.

Marketing in B2B technology is the part of GTM that creates the awareness and the demand the rest of the system converts. It is the element most commonly under-invested in early-stage scale-ups, the element most often mis-measured at established companies, and the element most likely to be blamed for problems that originate elsewhere in the funnel. Salamander’s marketing work usually runs through a CMO-as-a-Service seat, and it covers a set of interlocking areas: positioning, messaging, demand generation, brand, content, and the marketing-to-sales handover.

Positioning.

Positioning is the foundation. Who the company is for, what category it sits in, what alternative the customer is choosing it against, and why it is the better answer. Most B2B technology companies have positioning that has drifted — the company is selling to a different segment than its positioning addresses, or the category they pioneered has been redefined by competitors who arrived later, or the value proposition is still framed around what the product did three years ago. Sharpening positioning is the single highest-leverage piece of marketing work, and it has consequences that run through sales motion design, pricing, content, and recruitment. We push to look at positioning early on any marketing engagement.

Messaging.

Messaging is positioning translated into language the buyer can act on. The five-word version, the headline, the deck slide that opens the conversation. Messaging matters because it is the thing that has to travel — across the sales team, the website, the partner network, the inbound channel, and the customer’s internal champion as they sell the product up to their own leadership. Inconsistent messaging is one of the most common symptoms of unfinished positioning work. The work is to converge on a small number of statements that the company is confident enough in to put everywhere.

Demand generation.

Demand generation is the engine that produces the pipeline. The work spans the channels through which demand is created (paid search, paid social, organic search, content, events, partner-led demand, account-based motions) and the operating discipline that measures whether they are working. Most demand generation programmes we walk into have one of two problems: too many channels at sub-scale, with no single channel given enough investment to produce reliable returns; or one channel that has been the engine for so long that it has saturated and now needs to be supplemented before the pipeline starts to drop.

A note on attribution. Marketing attribution in B2B technology is genuinely hard — customer journeys are long, touchpoints are many, and the lift from any single channel is usually impossible to isolate cleanly. The Salamander posture is pragmatic. Use the attribution available, triangulate across multiple views, and resist both the temptation to over-claim and the temptation to dismiss demand generation as unmeasurable. The honest answer is usually “we have a directional read on what is working; we are running tests to sharpen it.”

Brand at B2B technology.

Brand in B2B technology is not the same thing as consumer brand, and treating it as if it were is one of the more expensive mistakes a scale-up can make. B2B brand is what the buyer believes about the company before the sales conversation begins — whether it is credible, whether it solves the buyer’s problem, whether it is safe to bring to an internal committee, whether other companies like the buyer’s own use it. It is built through customer outcomes, analyst recognition, thought leadership, employee advocacy, and the cumulative consistency of how the company shows up over many quarters. It is not built primarily through colour palette refreshes and tagline workshops.

Content.

Content is the medium through which positioning, messaging, and brand reach buyers across long sales cycles. The work spans the assets themselves (point-of-view pieces, case studies, demos, technical write-ups, analyst-grade material) and the operating discipline that produces them at the right cadence with the right quality. Most B2B technology companies we walk into produce too much content, too inconsistent, and in shapes that do not match how their buyers actually want to consume it. The Salamander work is rarely “produce more content” — it is more often “produce less, and produce it deliberately, around the assets the sales team and the partners will actually use.”

The marketing-to-sales handover.

The handover is the seam where marketing-generated demand becomes pipeline the sales team can convert. It is one of the most reliable sources of friction in B2B technology GTM. Marketing complains that sales is not working leads quickly enough; sales complains that the leads marketing produces are not qualified. Both complaints are often partly true. The work is to install the qualification criteria, the response time SLAs, the handover protocols, and the joint operating cadence that produce a clean handover regardless of who is right about which lead. A CMO-as-a-Service Associate is usually working alongside a CRO-as-a-Service Associate on this seam, and the engagements work better when both seats are filled.

Working with BAM on brand and campaign execution.

A CMO-as-a-Service seat covers the strategic shaping work this chapter has set out — positioning, messaging, the operating discipline of demand generation, the marketing-to-sales handover. It does not cover the full execution of a brand refresh or the build-out of a complete marketing campaign. Where a client engagement surfaces that kind of work, Salamander partners with BAM, a marketing agency we have a reciprocal referral relationship with. More on BAM at www.BAM.com.sg. BAM brings the team, the craft, and the throughput that brand and campaign execution require; Salamander brings the operator-level CMO seat and the commercial context. The two fit together cleanly.

The commercial arrangement is symmetrical. Salamander is incentivised to refer brand and campaign execution work to BAM in the same way BAM is incentivised to refer fractional CMO-as-a-Service and broader operating work to Salamander. When a marketing engagement surfaces an execution scope beyond what the CMO seat can carry, raising BAM with the client early is the right move. Loop in the partnership contact at Salamander to coordinate the introduction; do not run BAM in parallel to the Salamander engagement without that step.

Common failure modes.

Positioning that has not been sharpened in years, while the market has moved. The symptom is rising CAC and falling conversion; the cause is upstream.

Demand generation spread across too many channels at sub-scale. The fix is concentration — fewer channels, more invested in each, until each is producing measurable returns.

Brand work that prioritises aesthetics over substance. The new visual identity launches; the conversion rates do not move; the leadership team wonders why.

A marketing function that is producing content the sales team does not use. The fix is alignment around what sales actually needs to advance specific conversations with specific buyer personas.

Marketing and sales operating in parallel rather than as one system. The clearest symptom is a handover that does not work; the deeper symptom is that marketing and sales have not agreed on who the customer is.

For you as a Senior Associate working on a marketing engagement, the practical implication is to start with positioning and work outward. A marketing engagement that begins with channel optimisation, content production, or a brand refresh, without first checking whether the underlying positioning is sound, will produce more activity than impact. Sharpen the foundation first; let the rest follow.

Customer success is the part of GTM responsible for what happens after a customer has been acquired. The work spans the retention of the existing customer base, the expansion of revenue within it, the early-warning signals that flag customers at risk, and the recurring revenue mix that compounds when the function is doing its job. In B2B technology, where the lifetime value of a customer dwarfs the value of the initial sale, customer success is the function most directly tied to the long-term unit economics of the business.

This chapter sets out the substance.

What customer success actually does.

The motion looks different across the three GTM modalities introduced in Chapter 28. In a human-led enterprise motion, customer success usually runs through dedicated customer success managers attached to named accounts — onboarding the customer, running periodic business reviews, identifying expansion opportunities, surfacing risk signals to account managers, and renewing contracts when they come up. In a product-led motion, the customer success work runs partly through the product itself (in-product onboarding, usage prompts, activation flows) and partly through a smaller customer success team supporting the larger accounts where the product alone is not sufficient. In an e-commerce motion, the function is closer to consumer customer service plus retention marketing, and the design questions are about cohort retention rather than account management.

Across all three, the underlying logic is the same: customers acquired at the front of the funnel produce most of their value over time, and the function responsible for that time is customer success.

The LAER lens.

The work of customer success can be read through the four phases of the customer lifecycle that the LAER model captures. Land — the customer is acquired and the contract is signed. Adopt — the customer reaches working use of the product, the value the sale was anchored on is realised, and the relationship moves from “bought” to “used.” Expand — additional usage, seats, products, or premium tiers are sold into the account as the customer’s needs grow. Renew — the contract is renewed, or not, at the end of its term, on terms that reflect the value the customer has received. LAER is a widely-used industry frame for the post-sale lifecycle, and most Salamander GTM engagements that touch customer success use it explicitly as the structuring lens.

Two practical points follow. The first is that the four phases are not equally well-resourced in most B2B technology companies. Land is owned by sales and is usually well-instrumented and well-staffed; Renew is owned somewhere between sales and customer success and is typically the next-best resourced. Adopt and Expand are the phases where most companies under-invest — and they are also the phases that drive the unit economics that matter most. The assessment usually surfaces here.

The second is that the metric set we look at in the rest of this chapter maps onto LAER. Gross retention and renewal rates measure performance at Renew. Expansion economics, NRR, and the expansion motion measure performance at Expand. Customer health and time-to-value measure performance at Adopt. The first move on most customer success engagements is to set out the company’s performance against LAER, phase by phase, and identify which phase is the binding constraint.

Retention.

Two retention metrics matter. Gross revenue retention measures the portion of recurring revenue that stays in the customer base from one period to the next, before any expansion. Net revenue retention measures the same after expansion has been counted. A SaaS business with gross retention above ninety per cent and NRR above one hundred and ten is one where the existing customer base is compounding without any help from new acquisition; a business with gross retention in the high seventies and NRR below one hundred is one where new acquisition has to outrun an existing leak.

Most retention engagements at Salamander begin with the cohort view — retention by customer cohort, by segment, by product, and by acquisition channel — because the headline number almost always hides material variation underneath. A retention problem that looks general is usually a retention problem that is specific to a particular cohort or segment.

Expansion motion.

Expansion in B2B technology happens through several mechanisms: more seats or more usage of the existing product, additional products sold into the same customer, premium tier upgrades, and additional services. The motion that produces expansion is rarely the same motion that produced the original sale — it is closer in shape to ongoing customer management than to deal closing, and it relies on the customer success function understanding the customer’s evolving needs rather than on a salesperson re-prospecting the account.

Designing the expansion motion is part of the customer success work. Who is responsible for spotting expansion opportunities; how those opportunities are passed to the right closing role; what the cadence of expansion conversations is; how expansion economics are measured separately from new-business economics. Most B2B technology companies under-invest in this design and over-rely on individual customer success managers to figure it out account by account.

Customer health.

Customer health is the leading indicator that retention reporting cannot give you. By the time gross retention has moved, customers who were going to churn have already decided. The health signals that matter sit upstream of the renewal: changes in product usage, changes in the customer’s own business or leadership, shifts in support ticket patterns, changes in QBR engagement, the answers customers give to the right open questions in their relationship reviews.

Most companies we work with either have no formal customer health system, or have one that has not been calibrated against actual churn outcomes. Building or recalibrating this system is a recurring Salamander engagement scope. The discipline is to instrument the signals, to track them at the account level, and to make the customer success team act on them while there is still time to act.

Recurring revenue mix.

Customer success is the function most responsible for the trajectory of the recurring revenue mix that Chapter 21 set out as one of the most important economics in B2B technology. A customer success function that successfully expands customers, surfaces and resolves risk before it becomes churn, and renews contracts at acceptable price increases is the function that moves the mix in the right direction. A customer success function that does not do these things is, in operational terms, leaking the value of the customer base.

Common failure modes.

A few patterns recur.

Customer success treated as customer service. The function answers tickets and runs onboarding but does not own retention or expansion outcomes. The work is operationally narrower than it needs to be; the company sees the consequences in NRR.

Onboarding that under-delivers. The first ninety days of a customer’s life with the product set the trajectory for everything that follows. Companies with weak onboarding consistently see retention problems eighteen months later, when the cumulative weight of customers who never reached full value comes due.

Customer success that is not in the operating cadence. The function operates in parallel to sales, marketing, and product, with its own metrics that nobody else looks at. The signals that should be informing the rest of the GTM system stay siloed.

Expansion motion that depends on individual heroics. Two or three customer success managers produce most of the expansion revenue; the rest of the team renews at flat. The assessment is usually a motion design problem rather than a hiring problem.

Customer success ownership that has been left unclear. The CRO believes customer success belongs in sales; the COO believes it belongs in operations; the CEO has not resolved the question. The function ends up under-led regardless of where it sits.

For you as a Senior Associate working on a customer success engagement, the practical implication is to read three things first: the cohort retention picture, the customer health system (or its absence), and where customer success sits in the operating cadence. Those three usually tell you more about why the function is not producing than any amount of conversation about individual customer accounts.

The chapters in Part 5 so far have set out the substance of Salamander’s GTM discipline — the five elements that make up the system, the assessment that walks the system end to end, and what we do inside each element. This chapter is about the shapes those engagements take. As with Finance (Chapter 27), every GTM engagement is in some sense bespoke, but a small number of shapes recur. Knowing which one you are inside helps scope the work, calibrate the pace, and read what the client actually needs.

The list is not exhaustive. Patterns evolve with client demand, and new shapes appear. The three below are the most recurrent at the time of writing.

Shape 1 — Resolve growth issues, direct or via channel.

The most common GTM engagement shape. A B2B technology company has seen growth slow or stall, and is looking for senior operating capacity to assess the cause and fix it. The engagement typically opens with the assessment from Chapter 29 — walking the funnel end to end, identifying the binding constraint, naming it clearly — and then moves into the substantive work of addressing whichever element of the GTM system is producing the symptom.

These engagements usually run through a CRO-as-a-Service Associate, sometimes with a CMO-as-a-Service Associate alongside if the binding constraint sits in marketing or in the marketing-to-sales handover. The work is rarely contained to a single element of GTM. A direct sales motion that has stopped converting at the right rate may be the visible symptom, but the underlying cause could be ICP drift, marketing under-producing pipeline, channel conflict, or all three. The discipline is to fix the binding constraint first, see whether the symptom moves, and address the next one if it does not.

Engagements of this shape run anywhere from three months for a targeted intervention to two years or more for a full GTM rebuild. The pace varies with the urgency and with how much of the system needs reshaping.

Shape 2 — Improve channel profitability.

The second recurrent shape. A B2B technology company is leaning on a channel programme that is consuming margin without returning proportional revenue, or is failing to convert the channel partners it has signed into productive contributors. The engagement scope is narrower than the broader growth-resolution shape but often runs deeper into the channel mechanics covered in Chapter 31.

The work usually combines a productivity assessment (revenue and pipeline by partner, segment-level economics, programme cost-to-serve), a rationalisation of the partner base (concentrating on the partners that produce, tiering the next layer, letting the rest fall away), and a redesign of the partner economics and enablement so that the remaining partners can produce more. Because channel margin is, in unit-economic terms, a Finance question as much as a GTM one, this shape frequently runs alongside a Finance lens — particularly in engagements where the broader profitability picture is the driver of the work.

Shape 3 — Build a better marketing narrative.

The third recurrent shape. A B2B technology company has a positioning, messaging, or brand problem — usually some combination of the three — and the consequences are showing up in CAC, in win rates against specific competitors, in analyst coverage, or in the way investors describe the company to other investors. The engagement runs through a CMO-as-a-Service seat and goes deeper into the work set out in Chapter 33.

These engagements are sometimes commissioned ahead of a fundraise or a transaction process, when the gap between how the company sees itself and how the market reads it has become commercially expensive. They are also commissioned independently of any transaction, where a leadership team has recognised that the narrative has drifted out of alignment with what the company has actually become. Either way, the substance is the same: sharpen the positioning, converge on messaging, rebuild the content set that makes the story travel, and instrument the demand generation so the new narrative actually lands with the buyer.

Within this shape, Salamander also supports companies that are actively fundraising. The work extends beyond positioning and messaging into the materials investors will actually see — the teaser used to open the conversation, the pitch deck used through the process, and the sharpened investor narrative the leadership team will tell across the room. The CMO-as-a-Service seat usually leads this work, often alongside a CFO-as-a-Service Associate (Chapter 27) where the financial story needs to land at the same time as the strategic one. Where the work calls for execution-heavy design and production beyond what the CMO seat carries, BAM is the partner we bring in (Chapter 33).

Other shapes worth knowing.

The standalone fractional CRO-as-a-Service holding the seat while a permanent hire is recruited or developed. The standalone fractional CMO-as-a-Service playing the same role on the marketing side. The international growth engagement, covered in Chapter 32, which sometimes shows up as its own scoped piece of work and sometimes as a workstream inside a broader growth-resolution programme. The customer success rebuild, where the engagement is specifically focused on retention, NRR, and expansion motion (Chapter 34). The strategy-development engagement, where a client asks Salamander to develop or document a GTM strategy (or, occasionally, a total company strategy that spans more than GTM). The ask is not frequent, but it is in scope — see the related note in Chapter 12. The post-investment value creation engagement, often initiated through a PE relationship and described in Chapter 19.

A note on how shapes blur in practice.

GTM engagements blur even more readily than Finance ones, because the elements of GTM interact more tightly. A growth resolution engagement that begins with a sales assessment often surfaces a marketing problem, a customer success leak, or a channel issue that has to be addressed alongside. A channel profitability engagement that begins with partner rationalisation can reveal that the direct motion does not work, in which case the work has to expand into the direct side before the channel changes can stick. A marketing narrative engagement that lands a sharper positioning often surfaces the next question — whether the sales motion and customer success motion are calibrated to the customer the new narrative actually targets.

The shape tells you where you start. It rarely tells you where you finish, and the willingness to reshape the engagement as the work develops is part of the practice.

For you as a Senior Associate, the practical implication is the same as for Finance: name the shape on day one and keep naming it as the engagement evolves. A clear shared view of the shape — between you, the Associate leading the engagement, and the client — is what prevents scope drift later. When the shape genuinely changes, name that too, and reset the conversation rather than letting the engagement drift inside its original framing.

Part 6 covers the third of Salamander’s three disciplines. Operations at Salamander spans a broader functional scope than the word usually implies. It covers product development, product management, support services, managed services, professional services, and project management. Across that scope, the work we do clusters around four cross-cutting themes — delivery performance, managed services models, operating cadence at scale, and scalable execution. The chapters that follow take each theme in turn (Chapters 37 through 40) and close with the engagement shapes that recur in our Operations practice (Chapter 41). This opening chapter sets out what falls inside the discipline, where it connects to Finance and GTM, and the standard we hold the work to.

The functional scope.

The six functions Operations covers at Salamander sit on a continuum from product creation through delivery to ongoing support.

Product development is the engineering work that creates and evolves the product itself. The Operations remit at Salamander is not to direct technical decisions inside an engineering team but to ensure that product development capacity, priorities, and cost structure are aligned to what the business actually needs.

Product management is the function that decides what the product does next — what is built, what is shipped, what is deprecated. The Operations work here is usually about installing the discipline through which those decisions are made consistently and against commercial reality, rather than about taking the decisions themselves.

Support services is the function that handles customer-facing technical and product support — the L1, L2, and L3 motions where they exist, the tooling that runs them, the cost-to-serve they produce, and the customer experience they deliver.

Managed services is the function that operates a service on behalf of customers — running an outsourced environment, hosting a platform, managing a process — usually with software in the loop. As Chapter 20 set out, this is increasingly the centre of gravity for many of the B2B technology businesses we work with.

Professional services is the function that delivers project-based work to customers — implementations, customisations, consulting work that wraps around the product, complex onboarding. Where the company has a services-led software model, professional services is where a meaningful share of revenue and people sit.

Project management is the function that runs delivery to plan — internal programmes, customer-facing implementations, transformation initiatives. The Operations work here is usually about getting from a project management capability that exists in pockets to one that runs consistently across the business.

The four cross-cutting themes.

Across those six functions, our work clusters around four themes that the next four chapters take in turn.

Delivery performance. How a B2B technology business actually delivers what it has sold — across products, services, and the customer-facing functions that support both. Reliability of the service, productivity of the delivery teams, margins delivery produces, customer experience that follows. Delivery performance is where many of the unit economics of the business actually sit, and where many of the customer success problems addressed in Chapter 34 originate. Chapter 37.

Managed services models. Designing, standing up, or scaling the managed services models that an increasing share of B2B technology businesses now run. The migration from project-based delivery to managed services is one of the most consequential structural transitions a services-led business can make, and one of the engagement shapes most often produced in Salamander’s Operations practice. Chapter 38.

Operating cadence at scale. The meeting, reporting, and decision rhythms that hold a scaling business together — across the product cycle, the delivery cycle, and the support cycle. A leadership team operating without disciplined cadence cannot run a scaling business; a leadership team with disciplined cadence catches issues early enough to act on them. Chapter 39.

Scalable execution. Where to invest in process and where to leave it alone. Most scaling companies over-engineer some processes and under-engineer others; the work is to know which is which and to design the process discipline that fits the company’s stage and model. Chapter 40.

The fractional COO seat.

The role most of this work runs through is COO-as-a-Service. A Salamander Associate working as a fractional COO carries the operating leadership of the client for the share of the week the engagement requires — taking ownership of delivery performance across the six functional areas, running the operating cadence, presenting to the board, and shaping the priorities the broader Operations work depends on. Where a client is not at a stage that justifies a full fractional COO, the same Associate can be engaged against a narrower scope — one or two functions, or one of the four themes — without taking the operating seat itself.

Where Operations connects to the other two disciplines.

Operations is the discipline most directly tied to how unit economics actually behave in practice. Cost-to-serve, delivery margin, capacity utilisation, the cost structure of customer onboarding — these are Operations levers with direct consequences in the Finance discipline. The customer experience that delivery performance produces, and the time-to-value the onboarding motion creates, are direct inputs into the customer success motion covered in Chapter 34. A Salamander engagement that runs Operations in a silo is an engagement that will produce work that does not move the broader picture. The most valuable Operations engagements are the ones in which the work is in active conversation with Finance and GTM.

The standard we hold the work to.

Operations work at Salamander is held to the same standard the rest of the practice is. Delivery models we recommend have to survive the operating cadence they will run inside. Managed services models we design have to support the margin profile the Finance work depends on. Operating cadences we install have to produce decisions, not just attendance. Scalable execution work we recommend has to fit the company the business actually is, not the company a textbook might describe. The bar is the same one set out in Chapter 23 — investor-grade, decision-ready, and outcome-focused — translated into operational terms.

For you as a Senior Associate working in Operations engagements, the practical implication is to read each engagement against both the functional scope (which of the six functions the work most directly touches) and the thematic frame (which of the four themes the work clusters around). Most Operations engagements span more than one function and more than one theme, but rarely all six functions or all four themes. Naming the scope clearly is part of scoping the work well.

Delivery performance is the foundational theme in Operations work, because most of what the other three themes (managed services, operating cadence, scalable execution) ultimately depend on is whether the company can deliver what it has sold. A B2B technology business with strong commercial momentum and weak delivery performance produces unhappy customers, deteriorating retention, deteriorating margin, and eventually a reputational drag the GTM motion cannot outrun. Fixing delivery performance is among the highest-leverage work Salamander does.

This chapter sets out what delivery performance means in practice, how we assess it, and what we typically install.

What delivery performance covers.

In a B2B technology business, “delivery” extends across the six Operations functions introduced in Chapter 36. Product development delivers features and platform capability to customers. Product management delivers decisions about what the product does next. Professional services delivers implementations and customisations. Managed services delivers ongoing operation of a service. Support services delivers issue resolution and customer assistance. Project management delivers programmes to plan. Each is a delivery motion in its own right, and a delivery performance problem can sit in any of them — or, more commonly, in the seams between them.

Delivery performance therefore is not a single metric. It is the integrated outcome of several things behaving well together: reliability, productivity, margin, and customer experience.

The four dimensions.

Reliability is whether the service does what the customer expects, when it expects, at the quality it expects. For a SaaS product, this is uptime and feature stability. For a managed services business, it is service level adherence. For professional services, it is delivering implementations on time and in scope. For support, it is response time and resolution rate. Reliability is the most visible of the four dimensions — customers measure it directly.

Productivity is the rate at which the delivery functions produce output relative to the cost of producing it. Engineering throughput per developer; tickets resolved per support agent; project hours billed against project hours planned; managed services accounts handled per operations engineer. Productivity moves slowly and is rarely the first thing leadership pays attention to, but it is one of the most important inputs into margin.

Margin is the financial outcome of reliability and productivity together. Delivery margin by customer, by product line, by service offering, by professional services engagement. Most companies do not have visibility into delivery margin at the level required to act on it, and getting that view is often the first move in a delivery performance engagement.

Customer experience is what the customer takes away from the delivery they receive. NPS and CSAT capture some of it; the harder-to-measure signals — the customer’s sense of whether the company is in command of its own delivery, whether their account is being looked after, whether problems get resolved at the right level — capture the rest. Customer experience is the leading indicator of the retention work in Chapter 34.

Assessing delivery performance.

The assessment moves across the four dimensions in roughly this order. Start with the customer-facing signals — reliability data, customer experience scores, complaint patterns, the issues that are escalating into account management. Move to the productivity view — output per dollar in each of the six functions, where it has been improving, where it has stalled, where it has slipped. Then to margin — the cost-to-serve picture across customer segments, product lines, and service offerings. Finally, look at the seams — the handovers between product management and engineering, between engineering and support, between professional services and managed services, between any of these and the customer.

Most delivery performance problems sit in one of two places: a single function that is under-resourced or mis-designed for the work it is being asked to do, or a seam where two functions are handing work to each other without the discipline to do it cleanly. The assessment is to find which one.

Common failure modes.

A few patterns recur.

Engineering throughput that has not kept pace with the size of the company. Features ship late, bugs accumulate, technical debt rises, and the product roadmap loses credibility with the GTM team and the customer base. The assessment is usually not “hire more engineers” — it is more often about how engineering is organised, how decisions reach engineering, and how the work is prioritised.

Professional services that is unprofitable. Implementations run over budget, scope creep is not managed, the time recovered against the project plan is consistently below what was sold. The assessment usually involves both pricing and project management discipline together; fixing either alone rarely works.

Support that is overwhelmed. Volume has outgrown the team’s capacity, the L1/L2/L3 escalation discipline has broken, support is in firefighting mode, and the customer experience reflects it. The work is often to redesign the tiered support motion, instrument self-service and product fixes that reduce ticket volume, and bring the team back to a sustainable rhythm.

Managed services delivery producing inconsistent results. Different customer accounts receive different levels of service depending on which engineer is on rotation, which contract was signed, or which exception has accumulated. Standardising the managed services motion is one of the recurring scopes of work in this shape, and Chapter 38 covers it in detail.

Seams that drop work. The most common single failure mode. Product management hands a feature to engineering without sufficient context; engineering ships it to support without sufficient documentation; support hands a customer issue to managed services without sufficient handover; managed services flags an issue to account management without sufficient escalation. Each individual handover seems fine; the cumulative effect is that customer-facing delivery degrades.

Automation, AI, and shift-left.

Underneath the four dimensions, a structural shift is changing how delivery actually gets done. Three connected trends — automation, AI, and shift-left — are reshaping the cost structure of delivery, the size of the teams that produce it, and the level at which problems can be intercepted before they reach a customer. Salamander engagements that ignore this layer are assessing yesterday’s operating model.

Automation is the foundation. Build pipelines, test automation, infrastructure as code, provisioning automation, runbook automation, support workflow automation — the categories are well established and the technology mature. Most B2B technology businesses we work with have done some of this already, often without a coherent strategy. The work is usually to identify the manual work absorbing the most senior time or producing the most variation in outcomes, and prioritise automation against the cost-to-serve and reliability impact it produces. What is usually limiting is not the technology but the discipline to drive an automation programme to completion before the team moves on.

AI overlays automation and pushes it further. AI in engineering accelerates code generation, code review, and the management of technical debt. AI in support deflects routine tickets, summarises customer issues, suggests responses, and increasingly resolves a meaningful share of tier-one work without human involvement. AI in managed services drives anomaly detection, predictive remediation, and automation of operational decisions that previously required experienced engineers in the loop. AI in professional services accelerates implementation through configuration assistants and pre-built accelerators that reduce the bespoke work each engagement requires. The substantive AI conversation inside an Operations engagement is rarely “should we use AI” but “where, against what measurable change in cost, cycle time, reliability, or customer experience.”

Shift-left is the operating-philosophy companion. The principle is to move quality, security, testing, support, and customer-facing intervention earlier in the delivery cycle, where issues are cheaper to fix and where the customer never sees them. Shift-left testing into the engineering workflow rather than a downstream quality function. Shift-left security embedded into engineering rather than treated as a late-stage gate. Shift-left support through in-product help, self-service, and product fixes that resolve issue classes rather than individual incidents. Shift-left in omni-channel customer contact operations — where more of the resolution is deliberately handled at L1 (front-line agents, bots, self-service) and L2 (specialist support), so that L3, engineering, and other senior resources are reserved for the cases that genuinely need them. The economics here are direct: a ticket resolved at L1 costs a fraction of the same ticket resolved at L3, and the shift is aided by increasing levels of automation across the contact channels — with AI in particular collapsing work historically done at L2 down into the L1 tier. Shift-left customer success that catches the signals that precede churn before they become churn. The common thread is that the cost of resolving a delivery issue rises sharply the closer it gets to the customer (or the higher the tier the resolution requires), so moving resolution upstream produces both better reliability and better economics.

The Salamander work in this area is rarely about installing any one of these as a project. It is about reading the delivery operation, identifying where these levers produce measurable change against the four dimensions, and sequencing the work so that automation enables AI, AI enables shift-left, and shift-left compounds the gains across the system.

Installing the fix.

Delivery performance work usually combines instrumentation (getting the data visible at the level required to act), redesign of the function that is under-performing or the seam that is leaking, and installation of the operating cadence that holds the improvement in place. Chapter 39 covers the cadence piece. Without it, the instrumentation fades back into dashboards nobody reads and the redesign reverts within twelve months.

For you as a Senior Associate working on a delivery performance engagement, the practical implication is to think across all four dimensions and across all six functions. Reliability without productivity is unsustainable. Productivity without customer experience is invisible to the people paying for the service. The work is to read the integrated picture, find the binding constraint, and address it without losing sight of the others.

The migration from project-based delivery to managed services is one of the most consequential structural transitions a B2B technology business can make. For the company, it shifts the revenue base from one-off to recurring, changes the margin profile, reshapes customer relationships from transactional to durable, and changes how the business is read by investors and acquirers. For the team, it requires a different operating discipline than project-based delivery — one built around standardisation, repeatability, and continuous service of customers rather than the start-and-finish cadence of projects. This chapter sets out how we think about managed services and how Salamander engagements typically work with clients making the transition.

Why companies migrate.

The case for managed services usually combines four threads. Revenue mix improves — annuity revenue is more predictable, more durable, and valued more highly in a transaction. Margin behaves differently — at scale, a well-run managed services offering produces higher margin than project-based delivery because the cost-to-serve compounds down with each new customer rather than rising with each new engagement. Customer relationships deepen — managed services creates a continuous relationship rather than a periodic one, which improves retention and expansion in the customer success motion (Chapter 34). And the business reads more favourably to investors and acquirers, who value recurring revenue.

The migration is rarely free. It requires investment in tooling, automation, and standardisation that a project-based business has not previously made. It changes the sales motion. It changes the delivery economics in the first eighteen months before the long-term margin profile takes effect. The Salamander work is often to make this transition deliberately rather than reactively.

How we frame the work: Build. Sell. Deliver.

Salamander engagements in managed services often run against a three-phase frame we use to structure the work: Build, Sell, Deliver. Each phase is its own discipline, with its own deliverables and its own success criteria, and most managed services transitions run across all three.

Build is the work of designing and standing up the service. The service definition, the SLA structure, the underlying delivery motion, the tooling and automation that support it, the operating runbook, the pricing model. Build is mostly Operations work but has direct inputs from Finance (pricing, margin modelling) and from product (the technology architecture the service runs on).

Sell is the work of bringing the service to market. The sales motion calibrated to a managed services buyer rather than a project buyer. The positioning that distinguishes the new offering from the company’s project history. The contract templates that protect the standardisation of the underlying delivery. The pipeline cadence that tracks managed services pipeline separately from project pipeline. Sell is mostly GTM work but depends on the Build phase having produced something coherent enough to sell.

Deliver is the work of running the service for customers on an ongoing basis. The operating cadence that holds it together (Chapter 39). The customer success motion that retains and expands accounts (Chapter 34). The continuous improvement of the underlying service that protects the margin profile over time. Deliver is where the model lives or dies in practice, and where most of the long-term value of the managed services investment actually compounds.

Most of our managed services engagements involve some combination of the three. A company launching a new managed services offering needs all three in sequence. A company that has a service running but is failing to sell it needs Sell-phase work. A company that has been selling and delivering for years but cannot make the model produce margin needs Build-phase work — redesigning the underlying service — alongside Deliver-phase work to install the discipline.

What good managed services looks like.

A few elements characterise a managed services offering that works.

A clear service definition. What the customer is buying, what is included and what is not, what service levels apply, what happens when those levels are missed. Managed services that has not been deliberately defined will be defined in practice — by exceptions, scope creep, and ad hoc customer concessions that destroy the margin profile.

Service levels that mean something. SLAs that are measurable, monitored, and tied to commercial consequences. Companies that adopt SLAs as marketing language without instrumenting them rarely deliver against them, and customers notice. Companies that adopt fewer SLAs and meet them consistently build the trust that makes managed services durable.

Standardisation of the underlying delivery motion. The same operating procedures apply across the customer base. Different customers may have different commercial arrangements, but the underlying delivery should be largely standard. Variation in the delivery motion is the single most reliable way to destroy a managed services margin profile.

Pricing that reflects the value delivered, not the cost of delivery. The transition from cost-plus project pricing to value-based managed services pricing is a meaningful exercise in itself, and one Salamander engagements often run alongside the operational transition.

Automation and tooling that match the model. A managed services offering that depends on senior engineers manually executing the same tasks for each customer will not scale. Automation is not optional in this model; it is the architecture of the model. The themes in Chapter 37 — automation, AI, and shift-left — apply with particular force here. AI-driven anomaly detection, automated remediation, self-service customer portals, and tier-one ticket deflection are increasingly the difference between a managed services business that produces margin at scale and one that consumes it.

The shift from project-based to managed services.

Companies migrating from a project-based business model usually face three transitions running in parallel.

A commercial transition. The sales motion has to learn to sell a different shape of contract, against a different competitive set, with a different sales cycle and a different unit economics conversation with the buyer. Most sales teams that have sold projects for years need both retraining and partial restructuring to sell managed services well.

A delivery transition. The team that delivered projects has to learn to deliver continuous service. Engineers who measured themselves by project completion now have to measure themselves by sustained service quality. Project managers who handed off at completion now have to think in terms of ongoing customer accounts. This is a cultural transition as much as an operational one, and underestimating it is a common cause of stalled managed services migrations.

A financial transition. The revenue recognition changes. The margin profile changes — typically dipping in the first twelve to eighteen months as investment runs ahead of revenue, then recovering above the project-business baseline as the customer base accumulates. The cash profile changes. Boards and sponsors who are not prepared for the dip can lose confidence in the migration mid-way through. Part of the Finance and Operations work in a managed services transition is to model this dip honestly up front and manage the leadership team’s expectations against it.

Common failure modes.

A few patterns recur.

Managed services in name only. The company has rebranded its project work as “managed services” without redesigning the delivery model, pricing, or operating discipline. Customers see the inconsistency; margin does not improve; the migration produces optics without substance.

Under-investment in tooling and automation. The model is designed as if humans will scale linearly with customers. They will not. Without the automation and AI layer described in Chapter 37, managed services either consumes margin or fails to scale.

Standardisation that has been negotiated away. Each customer has been granted exceptions that, individually, looked reasonable. Cumulatively, they destroy the operating model. The fix is partly to honour the existing commitments and partly to install the commercial discipline that prevents new ones.

Customer success underestimated. A managed services business that does not invest in customer success treats every renewal as a fresh transaction. Retention suffers, expansion stalls, and the recurring revenue mix that justified the migration fails to compound.

For you as a Senior Associate working on a managed services engagement, two frames are worth holding side by side. Build, Sell, Deliver tells you which phase of the offering you are working on. Commercial, delivery, financial tells you which dimension of the organisational transition you are managing. A managed services migration that races ahead on the commercial side while the delivery side lags produces customer experience damage; one that races ahead on the delivery side while the commercial side lags fails to monetise the investment. The work is to sequence the phases and the dimensions so the company crosses the J-curve once, deliberately, rather than several times by accident.

Operating cadence is the third theme of Salamander’s Operations practice and one of the most under-appreciated. A scaling B2B technology business is not held together by individual heroics or by the strength of its product alone. It is held together by the meeting, reporting, and decision rhythms that let leadership see what is happening, decide what to do about it, and follow through. Companies with disciplined cadence outperform companies with the same talent and product running on improvisation, because the cadence is what turns intent into compounding execution.

This chapter sets out what operating cadence at scale looks like, how we install or refresh it, and why most scaling companies need our help with it more than they expect.

The cycles that matter.

Operating cadence at scale runs across several time horizons, each with its own purpose.

The weekly cycle. The cadence of operating reviews where the leadership team sees what is happening this week, against what was expected, with the decisions that need to be made before next week. Sales pipeline reviews, delivery reviews where managed services is in play, escalation forums for issues that cannot wait. The weekly cycle catches issues early enough to act on them.

The monthly cycle. The cadence at which the broader picture is reviewed — financial close, revenue and pipeline performance, delivery performance, customer health. The monthly business review is the heartbeat of a scaling company that runs well. It is also the cycle most often skipped or watered down when leadership teams are stretched, and it is the cycle whose absence most reliably correlates with companies that get into trouble.

The quarterly cycle. The cadence at which the company looks up rather than across — forecast revisions, priorities for the next quarter, where the year is heading. The quarterly cycle connects the operating reality of the business to the strategy and to the board. Done well, it is the rhythm in which most of a leadership team’s hardest decisions actually land.

The annual cycle. Planning, target-setting, leadership reviews, the conversations that shape the year ahead. The annual cycle is the slowest rhythm but the one that sets the parameters within which the other three cycles operate.

What good cadence actually involves.

A cadence that works has a few characteristics.

It produces decisions. The most common failure of an operating cadence is meetings that produce reports rather than decisions. Each meeting should have a small number of decisions named at the start, a small number reached by the end, and an owner attached to each. Meetings that consistently produce updates and nothing else are meetings that should be redesigned or removed.

It is attended by the right people. Operating reviews that include twenty people produce twenty-person conversations. The discipline is to invite the smallest set of people whose presence is required for the decisions on the agenda to be made, and to send notes to everyone else.

It is supported by the right data. The conversations only land where the data feeding them is trustworthy, on time, and at the right level of altitude. The Finance discipline of Chapter 24 — the close that closes on time, the rolling forecast, the KPI instrumentation — is the precondition for an operating cadence that actually works.

It is held to its rhythm. Cadences fade when the rituals slip. The monthly that becomes “every six weeks when we can fit it in” stops being a monthly. The weekly that is cancelled when something more pressing arises becomes a weekly that no longer holds the discipline it was designed to hold. The hardest part of operating cadence is not designing it but holding it through the periods when the leadership team would rather be doing something else.

Automation, AI, and shift-left in the cadence.

The operating cadence is one of the areas most directly reshaped by the trends introduced in Chapter 37. Automated dashboards that produce the operating data without manual rework. AI-generated summaries of the prior cycle that let leadership read the picture in a fraction of the time. Anomaly detection that flags issues before the operating review surfaces them. Self-service drill-downs that let function heads investigate without going back to Finance or BI for a custom analysis.

Shift-left applies at the cadence level too. The operating review that catches an issue before it requires escalation is a cheaper review than one that sees the same issue six weeks later, when remediation has become more expensive. Building a cadence in which signals are surfaced early — sometimes by AI ahead of human review — moves the locus of leadership attention from firefighting back to leadership.

Common failure modes.

A few patterns recur.

Cadence by accident. The company has meetings but no coherent cadence — different functions on different rhythms, board-facing reports that do not reconcile to internal reports, no clear ritual for the decisions that need to land each cycle. The work is to design the cadence deliberately and install the rituals to support it.

Cadence that has not scaled. The weekly conversation that worked at fifty people does not work at five hundred. The same room with the same people becomes a bottleneck rather than a forum. The fix is usually a layered cadence — function-level operating reviews feeding into a leadership-level review, with the layering itself designed against the company’s current size.

Cadence that produces optics rather than decisions. The operating review is a forum where the CEO is briefed on what everyone has been doing. Nothing gets decided. The fix is to redesign the agenda around decisions rather than updates, and to be willing to surface disagreements in the room rather than after it.

Cadence that has lost its discipline. The rhythm exists on paper but slips in practice. Monthlies get rescheduled, quarterlies get truncated, annual planning gets compressed into a week. The work is partly to redesign the cadence to be sustainable and partly to install the cultural expectation that the rhythm holds.

For you as a Senior Associate working on an operating cadence engagement, the practical implication is to read the existing rhythms first. What cycles exist, who attends, what decisions get made (and which ones do not), whether the data feeding them is sound. The assessment usually surfaces quickly. The substantive work — designing the layered cadence, installing the rituals, holding the leadership team accountable to them — is the part that takes the rest of the engagement.

Scalable execution is the fourth theme of Salamander’s Operations practice. It is the discipline of installing the right amount of process in the right places — enough that the business can scale without depending on individual heroics, not so much that the company sinks under its own administrative weight. The judgement of where the line sits is one of the recurring difficulties scaling companies face, and one of the more useful contributions Salamander makes.

This chapter sets out how we think about scalable execution, where process belongs and where it does not, and how we calibrate the discipline to the stage of the company we are working with.

The paradox.

Scaling a B2B technology business requires process. Without it, decisions get made differently by different people, customer-facing motions diverge, the same work gets done several ways depending on who is doing it, and the leadership team finds itself in the loop of decisions it should have stopped making years ago. With too much of it, the company moves slowly, capable people leave, and the agility that produced the early growth disappears under approvals and templates. The work is to install process where it earns its keep and to keep it out of where it does not.

Where process belongs.

Three categories of work benefit reliably from process discipline.

Repetitive work. Anything the company does the same way many times — customer onboarding, financial close, payroll, vendor management, code release, support escalation. Process here reduces variation, lowers cost-to-serve, and frees senior people from supervising work that should not need them. This is the category most directly reshaped by the automation and AI trends introduced in Chapter 37; the underlying repetitive work is also the work that automates most cleanly.

Cross-functional handovers. Anything that crosses a team boundary — marketing to sales, sales to delivery, delivery to support, product management to engineering. Process here is what stops work from being dropped at the seams, the failure mode identified in Chapter 37. The discipline is not bureaucracy; it is the shared expectation of what each side delivers to the other and on what timeline.

Customer-facing consistency. Anything the customer sees that should look the same across accounts — contracts, SLAs, communication standards, response times, brand presentation. Inconsistency at the customer-facing layer destroys trust faster than almost anything else, and process is what protects against it.

Where process does not belong.

A few categories do not benefit from process and are actively damaged by it.

Creative work. The early stages of product design, marketing narrative, strategic positioning. Forcing creative work through a heavy process pipeline produces predictable mediocrity. The work is to protect creative inputs from process before the right answer has emerged, and to apply process only at the point where the work has to be replicated or operationalised.

Edge-case judgement. The deal that does not fit the standard motion, the customer escalation that does not fit the standard playbook, the strategic question that does not fit the standard framework. These need senior judgement, not procedure. Companies that try to process-engineer every edge case end up with playbooks no one reads and decisions that take three times as long as they should.

Early experiments. The first attempt at a new market, a new channel, a new product. Process imposed before the experiment has been run produces premature commitment to a path that has not yet been validated. The discipline is to let experiments run, learn from them, and only then apply process to what works.

Calibrating process to stage.

The amount of process a company needs is a function of its size and complexity. A twenty-person scale-up does not need a vendor procurement process; it needs the CEO to make decisions on the few vendors that matter. A five-hundred-person scale-up does need one, because the alternative is that vendor decisions happen randomly across the business with no consistency. The same logic applies across most categories. The work is partly to know which stage a company is in, and partly to design process that fits that stage rather than borrowing it from a larger company the leadership team admires.

A useful heuristic: process should be installed when the absence of it has begun to produce visible cost — repeated work, inconsistency, errors, customer-facing damage — not before. Companies that install process speculatively tend to install too much; companies that wait until the cost is unmistakable tend to install it just in time.

Automation, AI, and shift-left in scalable execution.

The same trends introduced in Chapter 37 apply with particular force to scalable execution. Process that is automated is process that scales without consuming attention. Process that is AI-assisted — automated approvals against rules, anomaly flags inside operational flows, AI agents handling routine cases and escalating exceptions — scales further still. Shift-left applies to process design itself: the process that catches a problem upstream is cheaper than the process that catches it downstream.

The combination matters. The companies that scale execution well in the next several years will be the ones that automate the repetitive process, apply AI to the rule-based decisions, and reserve human judgement for the edge cases where it actually adds value. The Salamander work in scalable execution is increasingly to design with this layering in mind from the start, rather than to install human process that will then have to be reworked when automation catches up to it.

Common failure modes.

A few patterns recur.

Process imported from a larger company. The new VP joins from a Fortune 500, brings the process they ran there, and applies it to a company a fraction of the size. The result is bureaucracy that the company cannot afford and that the team will not follow.

Process built before the underlying work was understood. The procedure is documented for work that has not yet been done many times. When the work actually starts to happen, the procedure does not match it, and the team treats it as advisory at best.

Process that is documented but not enforced. Documents exist, no one follows them, no one notices that no one is following them. The cost is paid in the inconsistency the documentation was meant to prevent.

Process that has outlived the problem it was designed to solve. The approval workflow put in place to prevent a specific past error is still slowing down decisions years after the underlying risk has gone. The work is to remove process as deliberately as we install it.

For you as a Senior Associate working on a scalable execution engagement, the practical implication is to read the existing process inventory honestly. What is in place, what works, what is followed, what is documented but ignored, what is missing. The assessment usually surfaces both over-engineering and under-engineering in the same company. The work is to deliberately add where the absence is producing visible cost, deliberately remove where the presence is producing visible drag, and design the rest so it scales with automation and AI rather than requiring human enforcement at every step.

The chapters in Part 6 so far have set out the substance of Salamander’s Operations discipline — the six functional areas, the four cross-cutting themes, and what we do inside each. This chapter is about the shapes those engagements take. As with Finance (Chapter 27) and GTM (Chapter 35), every Operations engagement is bespoke in some sense, but a small number of shapes recur. Knowing which one you are inside helps scope the work, calibrate the pace, and read what the client actually needs.

The list is not exhaustive. Patterns evolve with client demand, and new shapes appear.

Shape 1 — Migrate to annuity-based service models.

The defining Operations engagement shape at Salamander, and one of the most consequential pieces of work we do. A B2B technology business is moving from a project-based or one-off revenue model toward a recurring managed services or annuity model, and is looking for senior operating capacity to design the transition and run it through. The engagement runs across the Build, Sell, and Deliver phases set out in Chapter 38, and across the commercial, delivery, and financial transitions that come with the model change.

These engagements cross-cut all three Salamander disciplines. The Operations work — the substantive design of the managed services offering and the operating motion behind it — runs alongside Finance work on the revenue recognition shift and the margin J-curve, and alongside GTM work on the sales motion calibrated to a managed services buyer. A Senior Associate on this kind of engagement should expect to be working with one or two other Salamander Associates carrying the Finance and GTM seats in parallel.

The engagement length varies considerably. A focused redesign of an existing managed services offering may run six to nine months. A full migration from a project-based business to a managed services model, with the operating motion and the commercial motion rebuilt around it, can run two years or longer. Boards and sponsors expecting the transition to land in a single quarter are usually expecting something the structural change cannot actually produce.

Other shapes worth knowing.

The standalone fractional COO-as-a-Service, holding the operating seat for a client while a permanent hire is recruited or developed, or where the operating role is genuinely fractional at the company’s current size. The substance of the work is whatever the operating role requires; the engagement is defined by the seat rather than by a particular workstream.

The delivery performance turnaround. A B2B technology business has a delivery quality problem, a delivery margin problem, or both, and the engagement is scoped specifically around the delivery performance themes in Chapter 37. The work usually combines an instrumentation phase, a redesign phase, and a sustainment phase, and it is one of the engagements where automation, AI, and shift-left are most likely to drive the substantive recommendations.

The operating cadence install. A scaling leadership team has reached the point where its existing rhythms have run out of useful life, and a Senior Associate is brought in to design and install the layered cadence the company now needs. These engagements are often shorter than the managed services migration but produce disproportionate value relative to their length, because the cadence is what holds everything else in place.

The process re-engineering or scalable execution programme. The client has either too much process slowing the business down, too little process producing visible cost, or both at once — and the engagement scope is to redesign the process inventory deliberately. Increasingly, this shape is also where AI and automation conversations are landing, because the most valuable process redesigns now anticipate which steps will be automated or AI-assisted within the next eighteen months.

The post-investment operations programme. A private equity sponsor has acquired a portfolio company and the value creation thesis depends on operating model change — margin improvement, delivery rebuild, managed services migration, or some combination. The engagement is often initiated through the PE relationship described in Chapter 19, and the operating partner shape Salamander sometimes plays inside a PE firm is one of the ways this work originates.

The carve-out or integration operations support. A transaction has happened — a divestiture, an acquisition, a corporate spin-out — and the operations function on one side or the other needs to be stood up, reshaped, or integrated. These engagements are time-bound around the transaction milestones and combine elements of several other shapes.

The AI and automation programme. Increasingly a standalone scope, where a leadership team has decided to move on AI adoption inside the operating model and needs senior operating capacity to design the programme honestly. The work is rarely “buy an AI tool”; it is more often about redesigning the delivery operation to take advantage of what AI and automation now make possible, against measurable change in cost, cycle time, and reliability.

A note on how shapes blur in practice.

Operations engagements blur for the same reasons Finance and GTM ones do: the work runs across the wider operating model, and a scope that opens as one shape often discovers a second or third inside it. A managed services migration that begins as a Build-phase engagement surfaces a delivery performance gap that has to be addressed before Sell can credibly run. An operating cadence install reveals that the underlying data is unreliable, and the Finance discipline work has to land before the cadence can hold. A delivery performance turnaround surfaces an automation and AI opportunity that becomes the substantive next phase of the engagement.

The shape tells you where you start. It rarely tells you where you finish, and the willingness to reshape the engagement as the work develops is part of the practice.

For you as a Senior Associate, the practical implication is the same as for Finance and GTM: name the shape on day one and keep naming it as the engagement evolves. A clear shared view of the shape — between you, the Associate leading the engagement, and the client — is what prevents scope drift later. When the shape genuinely changes, name that too, and reset the conversation rather than letting the engagement drift inside its original framing.

Hypothesis-driven problem solving is the first method in the Salamander toolkit because it sets up everything else. The discipline is simple to state and difficult to hold under pressure: in the first hours of a new problem, form an initial answer, and then run the rest of the work as a test of whether that answer is right. Refine, replace, or sharpen the hypothesis as the evidence arrives. Resist the instinct to gather everything before forming a view.

This chapter sets out why we work this way, how we do it, and what the first 48 hours of an engagement should look like.

Why hypothesis-driven.

The alternative is to begin with comprehensive analysis — collect all the data, talk to all the stakeholders, read all the documents, and let the answer emerge. The problem with that approach is that it produces a slow, low-momentum engagement, and it almost never converges on a useful answer in the time the client has. Hypothesis-driven work converges faster, because every piece of analysis is in service of a specific question, and because the team knows what would change the answer.

There is a second reason, particular to Salamander. Our engagements are fractional and time-compressed. The client is not paying for a six-month discovery phase; they are paying for senior operating capacity to move from analysis to decision to outcome at the pace of their business. Hypothesis-driven work is the method that fits that economics.

How to form a useful hypothesis quickly.

The first hypothesis on an engagement does not have to be right. It has to be specific, falsifiable, and consequential — specific enough to be tested, falsifiable enough that evidence can move it, and consequential enough that if it is right, the engagement has a path forward.

“The company has a growth problem” is not a hypothesis; it is a problem statement. “Growth has slowed because the sales motion was designed for a smaller deal size than the current ICP requires” is a hypothesis. The first version of the hypothesis on a Salamander engagement is usually formed in the first 24 to 48 hours, on the basis of the brief, the initial conversations with leadership, a first pass through the data the client has made available, and the pattern recognition the Senior Associate brings from operating experience.

The hypothesis is not yet the answer. It is the answer the team would commit to if no further information arrived. Holding it deliberately — knowing what we currently believe and why — is what makes the rest of the work focused.

Structuring the test.

A hypothesis is most useful when the test that would prove or disprove it has been named alongside it. For each significant claim in the hypothesis, the Associate should be able to state what evidence would change the view, what data is required to gather that evidence, and what the answer is likely to be if the evidence supports the hypothesis or contradicts it.

This produces a workplan that is sequenced by what would most decisively shift the hypothesis, not by what is easiest to investigate. The questions you would most want answered if you could only run three more analyses are the questions that should be run first. Everything that does not change the hypothesis can be deferred or skipped.

When the hypothesis is wrong.

Most initial hypotheses are partly wrong. That is not a problem with the method; it is the method working. The discipline is to update the hypothesis as evidence arrives, not to defend it. A Senior Associate who has formed a hypothesis on day three and is still defending the same hypothesis on day twenty, in the face of evidence that should have moved it, has stopped doing hypothesis-driven work and started doing post-hoc justification.

A useful internal signal: the hypothesis at the end of week two should be sharper, narrower, and more specific than the hypothesis at the end of week one. If it is not, either the evidence is not being read honestly or the work is not being done where the evidence is.

The first 48 hours of an engagement.

The first two days set the trajectory of the engagement. The shape of those days, in practice, looks like this. Read the brief. Have the first conversations with the engagement-defining leadership. Get a first pass at the data the client has made available. Hold an internal session — usually a working conversation between the Associate leading the engagement and the Senior Associate carrying the substantive work — to draft the initial hypothesis and the questions that would most decisively test it. Produce a short working memo, usually a single page, capturing the hypothesis, the supporting logic, and the workplan that follows. Share it with the client, or hold it for internal use depending on the engagement, and begin the substantive work.

Most engagements that go well at Salamander hit this rhythm in the first 48 hours. Most engagements that struggle skipped it and started with analysis without an answer in mind.

Common failure modes.

A few patterns recur. Hypotheses that are too general to test — “the company needs a better growth strategy” — and therefore impossible to disprove. Hypotheses formed on day one and never revised — held as conviction rather than as a working position. Hypothesis-driven work that has slipped, without anyone noticing, back into hypothesis-free data collection because the data has felt safer than the commitment to an answer.

For you as a Senior Associate, the practical implication is to form the initial hypothesis early and to hold it lightly. The first hypothesis is not the engagement’s deliverable; it is the engagement’s compass. The substantive work is to sharpen it through evidence. If you find yourself reluctant to commit to a hypothesis at the start of an engagement, the right answer is almost never to gather more data first — it is to commit to the best hypothesis you can form, name what would change it, and start the work.

Structuring problems is the second method in the Salamander toolkit and the natural companion to the hypothesis-driven work of Chapter 42. A hypothesis tells you what answer you currently believe; structure tells you what pieces of the problem the answer is built from, and which ones the evidence still needs to address. The two methods sit together: hypothesis without structure is a guess; structure without hypothesis is a workplan that does not converge.

This chapter sets out the structuring discipline as Salamander applies it, focusing on the two methods that do most of the work — the MECE principle and the issue tree — and on the judgement that turns a textbook structure into a useful one.

The MECE principle.

MECE — mutually exclusive, collectively exhaustive — is the test that distinguishes a useful structure from a misleading one. Applied to a set of categories or sub-problems, MECE asks two questions. Are the pieces non-overlapping, so that the same fact does not get counted in more than one place and the same decision does not have two homes? And do the pieces together cover the full ground, so that nothing important has been left outside the structure?

The discipline matters because most informal structures fail one or both tests. A list of “growth drivers” that includes pricing, marketing, sales, and customer success often has overlap (marketing influences pricing, sales feeds customer success) and gaps (where does competitive positioning sit?). Pulling the structure apart until it is mutually exclusive and putting it back together until it is collectively exhaustive is a deliberate exercise. It is also one of the most useful pieces of consulting craft that operators arriving at Salamander are most likely to underuse.

Issue trees.

The issue tree is the visible form of a MECE structure. A top-level question — usually the question the engagement is signed up to answer — is broken into the components that, taken together, would answer it. Each component is then broken into its own components, recursively, until the leaves of the tree are questions or hypotheses that can be tested directly with data and analysis.

A good issue tree has a few characteristics. It starts from a sharp, single question rather than a vague one. Each branch is MECE against its siblings. The depth of the tree matches the depth of the engagement — shallower trees for shorter engagements, deeper trees for substantive ones. And the leaves of the tree are testable, not just topical: “is the conversion rate in the mid-market segment lower than in the enterprise segment, and if so, by how much” is a leaf; “look at conversion” is not.

An illustrative tree for a growth-slowdown engagement:

Building one from a brief.

In practice, a useful issue tree on a Salamander engagement is built across the first week, not in one sitting. The first draft is sketched alongside the initial hypothesis from Chapter 42, often on a whiteboard or in a working memo. The first draft is almost never MECE; it has overlaps and gaps. The second draft, after a few conversations with the client and a first pass at the data, fixes most of the overlaps. The third draft, by the end of the first or second week, has been pruned to the branches that the workplan actually intends to investigate.

The tree is a working tool rather than a deliverable. The version that the client ever sees, if they see one at all, is usually a simplified version of the working tree that supports a particular conversation. The version the Salamander Associate maintains internally is sharper, deeper, and frequently updated. Treating the tree as a working artefact rather than a polished one is part of how it stays useful.

The relationship between structure and hypothesis.

Structure and hypothesis support each other. The hypothesis suggests which branches of the tree are most likely to contain the answer; the structure makes sure the branches the hypothesis is ignoring have not been overlooked by mistake. A well-formed hypothesis usually corresponds to a specific path through the issue tree — the path the team currently believes is most likely to lead to the answer.

This dual discipline produces a particular kind of workplan: focused on the branches the hypothesis points to, but with enough coverage of the other branches to ensure that if the hypothesis is wrong, the team will see it early rather than late. Engagements that have either hypothesis without structure or structure without hypothesis tend to converge poorly. The two together converge fast.

Common failure modes.

A few patterns recur.

Trees that are MECE in theory but not useful in practice. The categories are non-overlapping and exhaustive, but they do not map to where the answer actually lives. The fix is usually to redraw the tree around the hypothesis rather than from generic problem-decomposition templates.

Trees that are too deep or too shallow. A four-week engagement does not need a fifteen-leaf tree; an eighteen-month transformation programme needs more than three branches. Calibrating depth to engagement scope is part of the work.

Trees that have stopped being updated. The structure from week one is still on the wall in week six, but the engagement has moved past several of its branches and discovered two more that the original structure does not contain. A tree that is not being updated has stopped being a working tool.

Trees that exist only in one person’s head. One member of the team — usually the most senior — has the structure but has not shared it. The other Associates working on the engagement are inferring it from the tasks they are being given. This works briefly and then fails. Making the structure explicit and shared is part of how the team produces coherent work.

For you as a Senior Associate, the practical implication is to maintain a structure on every engagement you work, and to update it as the engagement moves. You do not have to invent it on your own; the Associate leading the engagement is usually carrying the working version. The expectation is that you understand it, contribute to it, push back on it where you see overlaps or gaps, and use it to anchor the analytical work you are doing. A Senior Associate who cannot describe the structure of the engagement they are working on, within thirty seconds, has not yet read the engagement carefully enough.

The third method in the Salamander toolkit is the use of assessment frameworks. Where hypothesis-driven work (Chapter 42) tells you to commit to an answer early and structuring (Chapter 43) tells you how to organise the problem, assessment frameworks tell you which lenses to reach for first when you walk into a B2B technology business and need to read it quickly.

An assessment framework is not a methodology to be applied dogmatically. It is a pre-built lens that accelerates the early read on a company, surfaces the patterns that experienced operators look for, and gives the team a shared starting point. The frameworks we use most are the ones that have been earned through many engagements — they have been refined, simplified, and pruned until they produce useful answers fast.

This chapter sets out how Salamander Associates use assessment frameworks, the categories of framework that recur in our work, and the discipline of knowing when to use them and when to set them aside.

The categories of assessment framework we use.

Three categories of framework do most of the work in early-stage engagement reads.

Discipline-level assessments. Each of Salamander’s three disciplines has its own working set. The Finance assessment from Chapter 24 — close discipline, rolling forecast, controls, KPIs, operating cadence — is the lens we apply first when we walk into a Finance engagement. The GTM assessment from Chapter 29 — walk the funnel from market to expansion, identify the binding constraint — is the lens we apply first when we walk into a GTM engagement. The Operations assessment in Chapter 37 — the four dimensions of delivery performance, plus the seams between the six functional areas — is the lens we apply first when we walk into an Operations engagement.

Business-model-level assessments. The model literacy from Chapter 20 functions as an assessment lens. The first read on any client is which model they actually run — pure SaaS, services-led, managed services, hybrid — and whether the model has drifted. The economics-that-matter framing from Chapter 21 is a related lens: the metric set worth attending to is different across models, and the early read includes picking the right set.

Stage-level assessments. The pressure points by stage from Chapter 22 — what tends to break at Series A versus the late scale-up versus the established business — is an assessment lens applied across every engagement. Reading the stage of the company is the move that tells you what to expect even before the data arrives.

The Salamander toolkit.

Salamander maintains a toolkit of a substantial number of PowerPoint decks covering recurring scenarios (Chapter 8). A meaningful share of that toolkit is built around assessment frameworks — the standard cuts of analysis we have found useful many times, the canonical exhibits, the visual frames for common questions. Before starting from a blank page on any assessment, an Associate’s first move is usually to check the toolkit. The work is faster, and the company benefits from the cumulative learning that has been refined into the toolkit over many engagements.

Alongside the deck toolkit, the firm maintains a current set of six assessment one-pagers — short A4 diagnostics designed to be used in the first two weeks of an engagement rather than presented. Three are discipline-level: the Finance Diagnostic, the GTM Diagnostic, and the Operations Diagnostic, mapping to Chapters 24, 29, and 37 respectively. Two are business-model-level: the Business Model Diagnostic and the Economics-That-Matter Diagnostic, mapping to Chapters 20 and 21. One is stage-level: the Stage Pressure-Points Diagnostic, mapping to Chapter 22. Each one-pager states the lens, the questions worth asking in the first two weeks, the binding constraints we see most often, the edges of the framework, and the output the assessment should produce. They sit deliberately outside the deck toolkit because their purpose is to accelerate thought, not to be shown to a client. All six are reproduced in full in Appendix A.

How the diagnostics sit alongside the playbooks.

The Salamander playbooks — the internal-use-only operational manuals introduced in Chapter 70 and listed on the dashboard described in Chapter 71 — each cover a specific engagement motion (SaaS go-to-market, building a managed services practice, customer success, pipeline management, channel strategy, and so on). Each playbook contains its own first-call diagnostic, designed as a forty-five-minute conversation with the executive sponsor once the motion is known, plus a set of capability plays that structure the engagement work that follows.

The six assessment frameworks in this chapter sit upstream of those playbooks. The handbook diagnostics tell an Associate which lens to read the company through; the binding constraint that read surfaces tells the Associate which playbook to open. The workflow runs in three steps: read the company through the relevant handbook diagnostic, identify the binding constraint, then route into the playbook (or playbooks) that own the motion that addresses it. Each diagnostic in Appendix A names its companion playbooks explicitly, and each named binding constraint points to the capability play inside that playbook where the work continues.

This relationship is also why the handbook diagnostics are deliberately lighter than the playbook first-call diagnostics. They are reading lenses across the business, not first-call scripts for a specific motion. An Associate uses them to converge on the right playbook quickly, and then uses the playbook to run the engagement.

How to use a framework well.

A framework is scaffolding for thought, not a replacement for it. Used well, a framework lets the Associate process the company faster, recognise patterns earlier, and converge on the binding constraint with less wasted analysis. Used poorly, a framework substitutes for thought — the Associate fills in the boxes, reaches for the standard conclusion, and produces an assessment that fits the framework rather than the company.

The discipline is to use the framework as a starting hypothesis about what to look at, not as a conclusion about what to find. Read the framework, read the company, and be willing to set the framework aside when the company does not fit it.

When to set the framework aside.

Frameworks have edges. The Finance discipline framework assumes a company at the size where the discipline is meaningful; it does not fit a forty-person scale-up the same way it fits a five-hundred-person business. The GTM funnel framework assumes a company that is selling at meaningful scale; it does not yet fit a company in pre-revenue product-led testing. The Operations framework assumes a company with delivery functions to assess; it does not fit a company whose product is entirely self-served and where Operations is essentially the product team.

The judgement to set a framework aside is part of the practice. A new Senior Associate who has been trained on the toolkit is sometimes inclined to over-apply it. A more useful instinct is to use the framework as long as it produces sharper questions and to abandon it the moment it starts producing answers that do not match the data in front of you.

Common failure modes.

A few patterns recur.

Framework as deliverable. The assessment exists in framework form — the chart, the matrix, the issue tree — but never converges on a recommendation. The framework has substituted for the answer. The fix is to remember that the framework is scaffolding and that the engagement is signed up to deliver something the framework can support but is not itself.

Framework mismatch. The framework being used does not fit the company being assessed. The Associate fills in the boxes anyway because the framework is what they know. The fix is to read the company more carefully and to be willing to design a framework that fits, or to set the framework aside entirely.

Framework drift. The framework starts the assessment, but the analysis quickly outgrows it without anyone updating the framing. The assessment becomes incoherent because the framework on the wall no longer reflects the work being done. The fix is to update the framing as the analysis develops, or to retire the framework deliberately.

Framework reverence. The Associate treats the framework as a complete account of the company rather than a starting lens. The result is an assessment that gets the framework right and the company wrong. The fix is the discipline that frameworks accelerate, rather than replace, senior judgement.

For you as a Senior Associate, the practical implication is to know the frameworks Salamander uses, reach for them early on engagements, but treat them as starting points rather than answers. The discipline of when to apply, when to adapt, and when to set aside is part of the craft this guide is trying to build. Most of that craft is learned by application rather than by reading; the guide can help you start, but the engagements are where the judgement actually gets developed.

Every recommendation rests on assumptions. Some are explicit and easy to test; many are implicit and rarely examined. The discipline of working with assumptions — making them visible, testing the ones that carry the most weight, and knowing which ones would change the answer if they failed — is the difference between a recommendation that survives contact with reality and one that does not.

This chapter completes the “defining the problem” section of the toolkit. The first three methods (hypothesis-driven work, structuring, assessment frameworks) tell you how to approach the problem. The fourth tells you how to be honest about what your approach is taking on faith.

The difference between a fact and an assumption.

A fact is something we know to be true on the evidence we have. An assumption is something we believe to be true without that evidence — sometimes because we have not gathered it yet, sometimes because the evidence is hard to find, sometimes because the assumption feels too obvious to question. The distinction matters because recommendations rarely fail on their facts; they fail on their assumptions.

Most assumptions on a Salamander engagement fall into a few categories. Assumptions about the market — the size of the segment we are targeting, the rate at which it is growing, the willingness-to-pay of the customers inside it. Assumptions about the company — the productivity of the sales team, the capacity of the delivery function, the rate at which a new hire will ramp. Assumptions about competitors — what they will do in response to a move, how their cost structure compares to ours, how customers will choose between them and us. Assumptions about timing — how long an implementation will take, when the next funding cycle is, when a customer renewal will close.

Making the assumptions explicit.

The first piece of work is to list the assumptions that the current recommendation rests on. In practice, this means walking the recommendation backwards — what would have to be true for this to work — and writing down every claim that has not yet been independently confirmed.

This step is more difficult than it sounds. Many of the most consequential assumptions on a recommendation are invisible to the team that built the recommendation, because they were absorbed as part of the company’s prevailing view. “The product will be ready in Q3” is sometimes a deliverable; it is more often an assumption. “The CRO will support this” is sometimes a confirmation; it is more often an assumption the team has not actually tested. The discipline is to write down the claims that look like facts but are actually assumptions.

Identifying the load-bearing ones.

Not all assumptions are equal. Some are decorative — if they are wrong, the recommendation shifts modestly. Others are load-bearing — if they are wrong, the recommendation does not stand. The work is to identify the load-bearing ones quickly and to focus the testing effort on them.

A useful question to surface load-bearing assumptions is: if this assumption is wrong, does the recommendation still work? If the answer is yes, the assumption is decorative. If the answer is no, the assumption is load-bearing and deserves the team’s serious attention. Most recommendations have a handful of load-bearing assumptions and a much longer list of decorative ones; focusing the testing effort on the handful is the discipline.

Testing the load-bearing assumptions.

Once identified, load-bearing assumptions should be tested directly. The tests vary by assumption type. Market assumptions get tested through customer conversations, data triangulation, and external benchmarks. Company assumptions get tested through internal data, function-head interviews, and a hard read on whether the team has actually delivered at the level the assumption requires. Competitor assumptions get tested through win/loss analysis, public information, and customer feedback. Timing assumptions get tested through the calendar and the recent track record of similar commitments.

The output of the testing is a sharper view of which assumptions hold, which do not, and which cannot yet be tested but should be flagged in the recommendation. Recommendations that name their untested load-bearing assumptions are more honest, more defensible, and more useful than recommendations that bury them.

Sensitivity and scenario thinking.

The companion discipline is sensitivity analysis: how much does the recommendation move if a particular assumption moves? A recommendation that depends on a single assumption holding tightly is a recommendation with concentrated risk; a recommendation that holds across a range of plausible assumption values is more durable.

In practice, the Salamander work is to model the load-bearing assumptions across a range — a base case, an upside case, a downside case — and to ensure the recommendation behaves acceptably in each. If the recommendation only works in the base case, the team needs to know that, and the client needs to know it too. This is the link to scenario thinking covered in Chapter 53.

Common failure modes.

A few patterns recur.

Assumptions absorbed as facts. The most common pattern. A claim has been repeated inside the company long enough that no one remembers it was once an assumption. The team builds on it without re-testing. The recommendation rests on a foundation no one has examined in years.

Load-bearing assumptions never tested. The team identifies the assumption but does not test it, usually because the test is uncomfortable or expensive. The recommendation goes out anyway, with the assumption treated as if it had been validated.

Sensitivity analysis that does not move the recommendation. The team produces a sensitivity table because process requires it, but the recommendation does not actually change across the range. The exercise becomes ritual rather than judgement.

Assumptions held inside one person’s head. The Senior Associate carrying the engagement knows the assumptions; the rest of the team is working as if the assumptions are facts. The result is uneven analysis, with parts of the work over-confident about claims that the Senior Associate would have flagged.

For you as a Senior Associate, the practical implication is to write the assumption list early and to keep it visible throughout the engagement. Every meaningful piece of analysis you produce sits on assumptions — yours and the client’s. Making them visible is part of the work, not optional. A useful discipline at the end of each working session is to update the assumption list with anything that surfaced in the conversation that was treated as fact but is actually an assumption. Most engagements that produce durable recommendations are engagements that maintained this discipline. Most engagements whose recommendations did not survive contact with reality skipped it.

Reasoning from evidence is the first method in the Analysis section of the toolkit. Where the four “defining the problem” methods told you what question to work on, the five “analysis” methods (Chapters 46 through 50) tell you how to work it. Reasoning from evidence is the foundation of the rest: the discipline of moving from what the data shows to what we are willing to conclude, honestly and with the right confidence.

This chapter sets out how Salamander Associates work with evidence — gathering it, triangulating across sources, calibrating confidence, flagging where judgement is doing the work, and knowing when we have enough to call it.

What counts as evidence.

Evidence is anything that materially shifts our confidence in a conclusion. In B2B technology work, the main categories are: quantitative data from the client’s systems (financial, operational, GTM); external data and benchmarks; conversations with the client’s leadership, team members, customers, and partners; and observations of how the business actually runs in operating reviews, working sessions, and day-to-day work. Each category has its strengths and its biases, and a recommendation that rests on only one category is more fragile than one that rests on several.

Triangulation.

Triangulation is the discipline of combining evidence from multiple sources to reach a conclusion that none of the sources would support on its own. The CFO says revenue is on plan; the sales pipeline data shows a coverage problem; three customer conversations surface a competitive issue not yet reflected in churn data. None of the three sources is sufficient. The combined view is.

In practice, the Salamander work is to make triangulation deliberate. For each meaningful claim in a recommendation, the question is: what is the evidence base? If the answer is a single source — particularly a single conversation, a single dashboard, or a single external report — the claim should be either confirmed against a second source or flagged as resting on a single point of evidence. Recommendations built on triangulated evidence survive scrutiny; recommendations built on single-source evidence often do not.

Calibrating confidence.

Not every conclusion is held at the same level of confidence. Some claims are well-supported, others are likely but not certain, others are working hypotheses still being tested. Calibrating which is which — and communicating that calibration to the client and to the Associate leading the engagement — is part of the work.

A useful internal vocabulary distinguishes between “we are confident that,” “the evidence suggests that,” “we currently believe that,” and “our working hypothesis is that.” Each phrase corresponds to a different level of confidence and a different level of risk if the underlying claim is wrong. Mixing the levels — treating a working hypothesis as a confident claim, or burying a confident claim inside hedged language — is one of the more common failures of analytical honesty.

Flagging judgement honestly.

Some claims in a recommendation are not derivable from evidence at all; they rest on judgement. The judgement of an experienced operator about how a leadership team will react to a recommendation. The judgement about how aggressively to compete in a new market. The judgement about whether the team can execute a particular plan.

These claims are still useful — judgement is part of what clients hire Salamander for. But they should be labelled as judgement rather than dressed up as evidence-based conclusions. The integrity value from Chapter 7 applies here. A recommendation that says “our judgement, based on our pattern recognition from similar situations, is that…” is more honest, and more defensible, than one that buries the same claim in language that suggests it has been derived from data it has not been derived from.

When data conflicts.

Data sources frequently conflict, particularly in larger companies where different functions instrument the same metric differently. Sales says pipeline coverage is 3x; Finance’s view of the same pipeline shows 2x. Customer success reports NRR of 108; the Finance close shows 102. The temptation is to pick the source that supports the working hypothesis. The discipline is to investigate the discrepancy and to understand which view of the underlying reality is correct before continuing.

Most conflicts have a reasonable explanation — different definitions, different time cuts, different inclusion rules. Surfacing the explanation is part of the assessment work, because the same conflict is often present in how the leadership team itself reads the business.

When you have enough to call it.

A recurring question in evidence work is when you have enough to commit to a conclusion. The honest answer is rarely “when we have everything we could possibly have”; it is “when additional evidence is unlikely to change the conclusion, and the cost of waiting is starting to exceed the cost of being marginally wrong.”

In practice, a useful heuristic is the converging hypothesis. If the third, fourth, and fifth pieces of evidence are all pointing to the same conclusion the second piece of evidence pointed to, additional evidence is unlikely to change the answer. If those pieces are pulling in different directions, more evidence — or a different kind of evidence — is required before a confident conclusion can be drawn.

Common failure modes.

A few patterns recur.

Single-source confidence. A claim rests on one data source or one conversation, and the team has treated it as confirmed. The work to surface this is to ask, for each significant claim, what the second source is.

Confirmation pattern. The team has formed a hypothesis and gathered evidence that confirms it, without seeking evidence that would disconfirm it. The fix is the discipline of asking what evidence, if it existed, would change the view, and then looking for it.

Hidden judgement. A claim that is doing the work of evidence-based reasoning is actually judgement. The team has not noticed, or has not labelled it. The fix is the explicit labelling described above.

Over-precision. The recommendation expresses a confidence the evidence does not support. The Salamander Associate has used confident language because the meeting required confident language, not because the underlying analysis warrants it. The fix is to calibrate the language to the evidence and be willing to say “we do not yet have enough to commit on this.”

For you as a Senior Associate, the practical implication is to ask, for every significant claim in the analysis you are producing, three questions. What evidence supports this? What other evidence has been considered, and how do the sources reconcile? What level of confidence is appropriate, and how should that confidence be communicated? An Associate who can answer these three questions for every claim in a recommendation is doing evidence-based reasoning at the standard the work requires.

The 80/20 instinct is one of the most consequential methods in the Salamander toolkit, and also one of the easiest to lose under engagement pressure. The principle is simple: most of the answer comes from a small share of the available analysis, and the discipline is to find that share early and concentrate the work there.

This chapter sets out what the 80/20 instinct looks like in practice on a Salamander engagement, where it tends to break down, and how to keep it operative through the substantive phase of the work.

The instinct stated plainly.

A B2B technology engagement has, in principle, an unbounded amount of analysis that could be done. Every metric could be decomposed further; every conversation could be extended; every model could be made more granular. The instinct is to ask, for any piece of analysis being considered: if this analysis produces its expected answer, does the recommendation move? If yes, the analysis is worth doing. If no, the analysis is decoration. The Salamander discipline is to do the work that moves the recommendation and to skip the work that does not.

This is more demanding than it sounds. The default of most analytical work, in operating companies as much as in advisory, is to spend effort where effort is comfortable — on the data that is most accessible, the questions that are most familiar, the analysis that is most expected. The 80/20 instinct cuts against that default. It asks the analyst to spend effort where the answer matters, even when that is also where the data is harder to find and the conclusion is harder to commit to.

Where the 80/20 tends to live.

A few patterns recur on Salamander engagements about where the high-leverage analysis actually sits.

On the binding constraint, not the symptom. The visible symptom of a GTM problem is usually pipeline; the binding constraint is often somewhere else (Chapter 29). Spending effort on pipeline analysis when the constraint sits in retention or ICP drift produces a thorough assessment of the wrong problem.

On the load-bearing assumption, not the decorative ones. As Chapter 45 set out, most recommendations have a handful of load-bearing assumptions and a long tail of decorative ones. The high-leverage testing effort sits on the handful.

On the customer-facing reality, not the management view. The leadership team’s view of the company is one input; the customer’s view is often a more telling one. Customer conversations on a sales engagement, customer health data on a retention engagement, customer feedback on a delivery engagement — these are usually where the 80/20 of the assessment insight lives.

On the operating reality, not the strategic plan. Many engagements include a strategy document that describes what the company intends to do. The 80/20 of the analytical insight is usually in what the company is actually doing, not what it intends. The gap between the two is often the engagement’s real subject.

Knowing when you have hit the 80.

The 80/20 instinct includes knowing when to stop. The conclusion at the end of week two is usually approximately the conclusion at the end of week six; the additional four weeks tend to sharpen it rather than change it. If they would change it, that is a signal that the analysis is still useful; if they will only sharpen language that is already clear, the additional time is decoration.

A useful internal check: if the engagement had to ship a recommendation today, what would it say? If the answer is approximately the same as what the team has been working toward, the engagement is at or past the 80; the remaining time should be spent on making the recommendation land rather than on sharpening it further.

When 80/20 does not apply.

A few situations require completeness rather than approximation. Investor-grade reporting in the run-up to a transaction has to be substantially complete; a buyer’s diligence team will look at every line. Regulatory and compliance work has to be complete by definition. Financial close discipline cannot afford 80/20 — the books either reconcile or they do not. The 80/20 instinct in these situations is not “do less analysis” but “spend the right effort where it matters and apply full rigour where the work has to be complete.”

The discipline is to know the difference between an analytical engagement, where 80/20 is the right posture, and a precision engagement, where it is not.

Common failure modes.

A few patterns recur.

Comfort-driven effort. The team has spent two weeks producing a thorough analysis of an issue that did not turn out to be load-bearing, while the load-bearing issue has received a single afternoon. The fix is to ask earlier and more often where the answer actually lives.

Completeness as a substitute for judgement. The team has produced exhaustive analysis on every branch of the issue tree because exhaustive analysis felt safer than committing to the branches that matter. The result is a thirty-page memo where a four-page memo would have produced a better recommendation faster.

False precision. The team has refined a number to three decimal places when the underlying confidence supports one. The precision misleads the reader about how solid the answer is.

The opposite failure — premature closure. The team has hit a quick conclusion and stopped before testing whether the 80 is actually the 80. Sometimes the most accessible answer is the wrong answer, and a little more digging would have revealed it.

For you as a Senior Associate, the practical implication is to ask the 80/20 question explicitly at multiple points in the engagement: at the workplan stage, at the first analytical review, at the midpoint, and at the final synthesis. The honest answers to “what is producing the answer” and “what is producing decoration” should keep shifting through the engagement. When they stop shifting, that is usually the signal that the engagement has converged.

Data analysis on a Salamander engagement runs at a different altitude than data analysis inside an operating function. The function head’s job is to know the data deeply, run reports against it, and operate the metric system day to day. The Senior Associate’s job is to read the data at the altitude where the recommendation is being made — sharply enough to see the discontinuities, deeply enough to understand what is producing them, but not so deeply that the senior judgement gets buried under the granularity.

This chapter sets out how Salamander Associates approach data analysis, the techniques that matter most, and the discipline of staying at the right altitude.

The altitude question.

Most data on a B2B technology engagement can be analysed in detail that goes far beyond what the recommendation requires. A salesperson’s individual conversion rate is interesting if the question is whether to keep that salesperson; it is decoration if the question is whether the overall sales motion is working at scale. The Senior Associate’s altitude is set by the question the recommendation has to answer. If the question is at the company level, the analysis stays at the company level until something at the function level is unmistakably different from the average. If the question is at the function level, the analysis goes one level deeper.

A useful internal check: am I producing analysis a function head could not produce, or am I duplicating analysis they already have? The former is where the value sits.

Cohort analysis.

Cohort analysis is the single most useful technique in B2B technology work, and one of the most under-used by leadership teams. It groups customers, employees, products, or whatever the unit is by their starting time period and tracks each cohort over time. The point is that the headline number almost always hides differences between cohorts that matter for the assessment.

A SaaS business with overall NRR of 105 might have first-year cohorts at 120 and third-year cohorts at 90. The headline is fine; the trajectory is not. The assessment question — what is happening to retention as the customer base ages — is invisible without the cohort cut. The same applies to customer success: an aggregate retention number that has been stable for two years can hide deteriorating retention in the most recent cohorts that will become the aggregate problem in eighteen months.

Cohorts are useful wherever a recurring relationship exists — customer cohorts, employee cohorts, product cohorts. The discipline is to ask, for any meaningful aggregate metric, what cohorts produced it, and whether the trajectory across cohorts tells a different story than the aggregate.

Segmentation.

Segmentation is the related technique of slicing the data by meaningful categories — customer segment, geography, product line, channel, deal size, industry. The aggregate hides variation; the segmented view often shows the variation more clearly than the cohort view does, particularly for cross-sectional questions rather than time-series ones.

The discipline in segmentation is to pick segments that map to where the answer might live, not segments that are easy to produce. A “customer size” segmentation that splits the base into small, medium, and large might be useful, or might mask the variation if the meaningful split is by buyer persona rather than by size. Choosing the right segmentation is part of the analytical judgement.

What to look at first.

A useful sequence on most B2B technology engagements is to walk the data in three passes.

The aggregate pass. Headline metrics over time. What is growing, what is shrinking, what is stable, what looks unusual against expectation.

The cohort pass. The same metrics broken by relevant cohorts. Where do the trajectories diverge from the aggregate? Where is the headline number masking a problem in a specific cohort?

The segment pass. The same metrics broken by relevant segments. Where is performance concentrated? Where are the outliers? Is the business actually one business, or three businesses with very different economics being managed as one?

By the end of the three passes, the Senior Associate should be able to describe what the data is showing in three or four sentences, with the discontinuities named, the patterns identified, and the questions for deeper analysis prioritised.

What to leave alone.

Not every data set deserves deep analysis. Some are too small to be meaningful, some are too noisy to support a conclusion, some are interesting but not connected to the recommendation. The discipline is to leave these alone. A Senior Associate who runs a thorough analysis on every available metric, including the ones that do not change the recommendation, has not yet applied the 80/20 instinct from Chapter 47 to their analytical work.

Common failure modes.

A few patterns recur.

Analysis at the wrong altitude. The Senior Associate has gone three levels deeper than the recommendation requires, and now cannot summarise the answer at the level the client needs to hear it.

Cohort or segment analysis run on too small a base. Cohorts of fewer than ten customers, segments of fewer than five accounts — the variance overwhelms the signal. The fix is to combine until the bases are meaningful, or to be honest that the data does not yet support the cut.

Headline numbers accepted without cohort or segment checks. The aggregate looks fine, and the team has not asked whether the cohorts and segments tell a different story. The retention problem two years away has not been surfaced.

Analysis without comparison. A number on its own — pipeline coverage of 2.3x, NRR of 104, gross margin of 68 — is not yet a conclusion. Comparison against time, against plan, against benchmarks, against cohorts is what turns the number into a finding.

Analysis without “so what.” The analysis has been done; the finding has been named; the implication for the recommendation has not been stated. The work is incomplete until the implication is on the page. Chapter 54 returns to this point.

For you as a Senior Associate, the practical implication is to set the altitude of every piece of analysis you produce against the recommendation it has to support, and to walk every data set through the aggregate-cohort-segment passes before going deeper. An Associate who can describe what the data is showing in three or four sentences, with the discontinuities named, is doing senior-altitude analysis. An Associate who cannot — who has to walk the client through twenty exhibits to make a point — has not yet found the altitude.

Models — the spreadsheets, simulations, and structured calculations Salamander Associates use to test recommendations and quantify their implications — are one of the most useful tools in the Analysis section of the toolkit, and one of the easiest to misuse. A model is a working hypothesis about how part of the business behaves. Built well, it sharpens the recommendation; built poorly, or trusted too far, it produces false precision and false confidence.

This chapter sets out how Salamander Associates build models, the disciplines that matter most, and how to keep a model useful through the engagement rather than letting it become a substitute for judgement.

The kinds of model we build.

Most Salamander engagements involve at least one of three model categories.

Financial models. Revenue, margin, cash, and the operating P&L over time. The CFO-as-a-Service work in Part 4 frequently produces a rolling forecast or a scenario model that becomes the working operating model for the company. Transaction-readiness engagements often involve a model the buyer or investor will scrutinise.

GTM models. Pipeline coverage, sales productivity, channel economics, customer success motion, retention and expansion behaviour. The CRO-as-a-Service work in Part 5 frequently produces models that read what the GTM engine is doing now and what it would do under a different design.

Operating models. Delivery capacity, cost-to-serve, managed services unit economics, capacity utilisation, the resource-and-margin profile of an operating function. The COO-as-a-Service work in Part 6 frequently produces these.

In practice, the most useful models for a B2B technology company combine elements of all three — a financial model that connects to a GTM model that connects to an operating model, so that a change in any of them propagates through to the others.

The Salamander model library, and a note on AI.

Salamander maintains an extensive library of Excel models and templates, mainly financial, built up over many engagements. Senior Associates joining Salamander often start by reviewing what is already in the library before building anything from scratch.

A meaningful share of the modelling work the library was built to support has been surpassed by what generative AI now makes possible — faster iteration, more flexible structure, lower cost of producing a custom model for a specific situation rather than retro-fitting a legacy template. The library is still useful where the structure is sound and the assumptions are well-tested; it should be set aside where rebuilding from scratch with AI assistance produces a better, more flexible answer faster. The judgement of which is which is part of the practice.

Building lean.

The discipline in building a model is to make it as simple as the question it has to answer requires, and no simpler. Over-engineered models — hundreds of inputs, dozens of sheets, complex interconnections — produce false precision and become impossible to maintain. Under-engineered models miss the dynamics that actually matter.

A useful principle is the minimum viable model. What are the inputs that materially change the answer, what are the outputs the recommendation depends on, and what is the simplest structure that connects the two? Anything beyond that is added later if and when it earns its place. Most Salamander models we walk into are too complex rather than too simple. The first piece of work on them is often to strip back.

Driver-based.

A model that does well is driver-based. The outputs flow from a small number of structural drivers — pricing, volume, mix, productivity, retention, capacity — rather than from a long list of hard-coded numbers. Driver-based models let the user ask “what happens if this driver changes” and see the answer cleanly. Hard-coded models let the user change a number and see what happens to the spreadsheet, which is not the same thing.

Identifying the right drivers is part of the work. The drivers should be the variables that genuinely move the business and that the leadership team has the agency to influence. A driver-based model on a B2B technology company usually has between five and fifteen meaningful drivers; everything else is downstream.

Stress testing.

A model is most useful when it has been stress tested against the conditions under which the recommendation might fail. The standard moves are sensitivity analysis on each driver, scenario analysis with combined movements, and downside cases that test how the recommendation behaves when several drivers move adversely at once.

The output of stress testing is a sharper view of which drivers matter most, where the recommendation is most exposed, and what would have to be true for the recommendation to break. The integrity value from Chapter 7 applies here: if the model produces a recommendation that only works in the base case, the team needs to say so honestly rather than burying the dependency.

Not falling in love with the model.

The most common failure of model work is over-attachment. A team has built a model carefully, the model produces an answer, and the team begins to defend the model rather than the underlying business judgement. Models are working tools — they exist to support the recommendation, not to be the recommendation. The discipline is to use the model where it sharpens the answer and to set it aside where it does not.

A useful internal signal: would the recommendation change if the model produced a different output? If yes, the model is doing the work. If no, the model is decoration — the recommendation rests on something else, and the model is being produced as supporting paperwork rather than as the genuine basis.

The relationship between model and recommendation.

A model that supports a Salamander recommendation should pass a few tests. It should be the right size — complex enough to capture the dynamics that matter, simple enough that another Associate can pick it up and understand it in an hour. It should be driver-based. It should have been stress tested. And the recommendation it supports should be derivable from the model — not derived elsewhere and then back-fitted to the model.

The last point is the one most often missed. A recommendation that has been formed in a working memo and then sent to a model for “supporting analysis” is doing model work backwards. The Salamander discipline is to develop the recommendation and the model in conversation with each other, so that the model is genuinely shaping the answer rather than dressing up an answer that has already been decided.

Common failure modes.

A few patterns recur.

The model that nobody else can read. Built by one Associate, with idiosyncratic structure, undocumented assumptions, and inconsistent logic. Useful only to its author. The fix is to build with the second reader in mind from the start.

The model that has outgrown its question. Started small, accreted complexity over many engagement weeks, now has hundreds of inputs that no one is actually using. The fix is to prune deliberately — most over-engineered models can be cut by half without losing useful precision.

The model that is hard-coded. The numbers feed forward but cannot be easily changed; sensitivity analysis is painful; scenario testing is barely possible. The fix is to refactor toward driver-based structure before the engagement needs the flexibility.

The model that the team has stopped questioning. The output is taken as authoritative because the model produces it; the underlying drivers are no longer being challenged. The fix is the regular discipline of asking “what would we have to believe for this model to be wrong, and is that what we believe.”

For you as a Senior Associate, the practical implication is to treat every model as a working tool, not a deliverable. Build lean. Build driver-based. Stress test it. Be willing to throw it away when the recommendation goes a different direction than the model first suggested. The model is in service of the answer; the answer is not in service of the model.

Quantification is the work of turning qualitative findings into numbers — the size of an opportunity, the magnitude of a risk, the impact of a recommendation. It is the analytical step that turns “this matters” into “this matters by approximately this much,” and that turns “this is a concern” into “this is a concern of approximately this scale.” Without it, recommendations remain vague; with it, the client can make decisions calibrated to consequence.

This chapter sets out how Salamander Associates approach quantification, the techniques that recur in our work, and the discipline of being honestly precise when the underlying data does not support exact answers.

Why quantification matters.

Numbers force precision. A recommendation to “improve margin” is impossible to act on; a recommendation to “improve margin by approximately three to five percentage points over eighteen months, with roughly seventy per cent coming from pricing and the remainder from cost-to-serve” gives a leadership team something concrete to plan against. The same logic applies to risks. “There is a competitive risk” is a sentence on a slide; “there is a competitive risk worth approximately twelve to eighteen months of growth if the competitor moves on price” is a risk the leadership team can prepare for.

Numbers also enable trade-offs. A leadership team faced with three potential moves cannot compare them sensibly without a sense of the prize each is going after. Sizing each one credibly is what lets the team decide.

Sizing a prize.

Sizing the prize of an opportunity is the most common quantification task on a Salamander engagement. The discipline is to land on a number — usually a range — that is defensible against an experienced challenger.

Two approaches do most of the work, and the most credible sizing uses both.

Top-down sizing. Start from a large external number — total market size, addressable customer base, revenue pool — and work down through filters that narrow it to the relevant scope. The advantage is breadth; the risk is that the filters multiply and the final number is more an artefact of the assumptions than the underlying reality.

Bottom-up sizing. Start from a unit — a customer, a transaction, an account — and multiply by the relevant base. The advantage is groundedness; the risk is that the unit assumed is not representative of the population.

Triangulation between the two is the discipline. If a top-down estimate and a bottom-up estimate produce numbers within plausible distance of each other, the sizing is more defensible. If they produce numbers far apart, one or both has a hidden problem that has to be surfaced before the number is shared.

Sizing a risk.

Sizing a risk is the mirror image. The discipline is to translate a qualitative concern into a quantified range the leadership team can weigh against other concerns and other opportunities.

A useful frame is to ask three questions. What is the probability the risk materialises in some form? If it does, what is the likely magnitude — revenue at risk, margin compression, cash impact, time lost? And what is the timing — over what period would the consequence land?

Combining the three produces an expected-impact estimate that lets the risk be compared against others. It also helps surface the cases where a low-probability, high-magnitude risk deserves attention out of proportion to its expected value — the asymmetric downside that has to be addressed even though it may not materialise.

Sizing with imperfect data.

Most of the quantification work on a Salamander engagement is done with incomplete information. The temptation is to wait until the data is better. The discipline is to size with the data available, name the uncertainty honestly, and refine as better data arrives.

A useful internal vocabulary distinguishes between a “directional” estimate, an “indicative” estimate, and a “committed” estimate. A directional estimate says which way the answer leans and roughly at what scale; it is appropriate early in an engagement. An indicative estimate says the answer is in a particular range, with the range’s width reflecting the underlying uncertainty; it is appropriate when the assessment is converging. A committed estimate is a number the team would stand behind in a board meeting; it is appropriate at the end of the engagement, after the work has been done.

Mixing the three — presenting a directional estimate with committed-estimate language, or hedging a committed estimate as if it were merely indicative — is one of the more common failures of analytical honesty.

Honest precision.

A number presented to two decimal places implies a confidence the underlying data rarely supports. A number presented as a round figure or as a range honestly signals the underlying uncertainty. The Salamander discipline is to use the precision level that the data and the analysis actually support — and to be willing to say “between three and five per cent” rather than “3.7 per cent” when the data does not support the third significant figure.

Clients and boards generally respect honest precision more than false precision. Buyers and investors are even less tolerant of false precision; they will find the over-claim, and the cost of being caught is much higher than the cost of having said “between three and five per cent” in the first place.

Common failure modes.

A few patterns recur.

Sizing that has not been triangulated. A single approach has produced a single number, and the number has been used without checking it against another approach. The fix is to do the second approach, even quickly, before relying on the number.

Sizing built on a single assumption. A top-down estimate that depends on a market-size figure from a single third-party report; a bottom-up estimate that depends on a single client interview. The fix is to test the assumption explicitly and to widen the range if it is not yet defensible.

Precision that exceeds confidence. A number presented to two decimal places when the underlying confidence supports one. The fix is to round.

Sizing language that hides uncertainty. The number is presented as a fact when it is actually an estimate. The fix is the calibrated vocabulary above.

For you as a Senior Associate, the practical implication is to bring a number to every qualitative claim you make in an engagement. The number will not always be precise — it cannot always be — but it should be honestly held, triangulated where possible, and calibrated to the confidence the data supports. The discipline of quantification turns advisory work from suggestion into decision support, and decision support is what Salamander engagements are signed up to deliver.

Synthesis is the craft of turning a pile of interviews, data, and observations into a structured answer. By the middle of a Salamander engagement, the team usually has more raw material than it can carry in any single head — meeting notes, data cuts, models, half-formed ideas, partial findings, side observations from the operating reviews. Synthesis is the work of converting that pile into a recommendation a client can act on.

This chapter sets out how Salamander Associates do that synthesis, the techniques that recur, and the discipline of producing a clear answer from material that is rarely clear on its own.

What synthesis is, and what it is not.

Synthesis is not summary. Summary reduces; synthesis produces. A summary tells the reader what was found; a synthesis tells the reader what to do about it, with the supporting argument structured so that the conclusion is the load-bearing element rather than the appendix.

Synthesis is not the same as analysis either. Analysis takes a question and produces a finding; synthesis takes the findings and produces a recommendation. Both are necessary; either alone is incomplete. A Salamander engagement that has produced a lot of analysis but has not yet synthesised has not yet earned its keep.

The practical methods.

Three methods do most of the synthesis work on a Salamander engagement.

Theme extraction. Walk through every meaningful piece of evidence the engagement has produced — interview notes, data cuts, observations — and group what they say into themes. The themes should be MECE in the structuring sense (Chapter 43): non-overlapping and collectively covering the material. Most engagements at synthesis time have between three and seven dominant themes; if the team is producing twenty, the themes have not yet been reduced enough.

Theme weighting. Not every theme matters equally. Some themes are central to the recommendation; some are peripheral observations that do not change the answer. Weighting the themes against the recommendation they support is part of synthesis. The peripheral themes get held to one side; the central themes carry the body of the work.

The single-sentence answer. The cleanest synthesis test is to ask whether the team can state the recommendation in a single sentence. Not the full argument; the answer. If the single sentence cannot be produced, synthesis is not yet complete. Most engagements that go well at Salamander reach a single-sentence answer by the middle of the engagement, with the rest of the work used to test, sharpen, and support it.

From observations to themes.

The move from observations to themes is the step most often skipped. The team has fifty bullet points of finding-level material; the temptation is to organise them into a deck and call it synthesis. The discipline is to ask, for each bullet point, what underlying theme it belongs to, and then to look at the themes rather than the bullets.

A useful internal move is to write each finding on a card (physical or virtual) and sort them. The themes emerge from the sort. The themes that emerge are usually fewer and sharper than the team expected.

From themes to answer.

The move from themes to answer is the second step most often skipped. The team has consolidated the findings into clear themes; the recommendation is then assumed to follow from the themes. It rarely does automatically. The themes describe what is true about the company; the answer is what to do about it. The bridge between the two is judgement, not arithmetic.

The judgement involves trade-offs that the themes themselves cannot make. Which of the themes is most actionable? Which produces the largest commercial impact if addressed? Which has the leadership team’s support already; which has to be won? Which fits the company’s stage and capacity; which would be right for a different company at a different stage? The synthesis is the place where these trade-offs land.

Building iteratively.

Synthesis is not done in one sitting. The first version of the answer is sketched early in the engagement, often as the initial hypothesis from Chapter 42. As the analysis develops, the synthesis is revisited and sharpened. By the middle of the engagement, the synthesis should be visibly tighter than it was at the start. By the end, it should be sharp enough to be defended against challenge.

A useful practice is to write the synthesis page (or the storyline) early — sometimes in week two or three — and to update it explicitly as the work develops. Engagements that wait until the end to write the synthesis often discover, too late, that they cannot.

Common failure modes.

A few patterns recur.

Findings without synthesis. The deck or memo lists what was found but does not say what to do. The reader has to do the synthesis themselves, which is the work the engagement was supposed to deliver.

Themes that are too many. The team has produced twelve themes when three would have served. The reader cannot hold twelve themes in mind; the synthesis is effectively a list rather than an argument.

Themes that are too neat. The team has produced themes that fit the framework cleanly but do not capture the messiness of the actual material. The synthesis reads well but does not survive contact with the client’s lived reality.

Synthesis that arrives only at the end. The team has done the analysis but has not built the synthesis through the engagement. The recommendation has to be assembled in the final week and does not have time to be tested.

For you as a Senior Associate, the practical implication is to synthesise continuously throughout the engagement, not at the end. Write the single-sentence answer early, even if it is wrong. Group findings into themes as they arrive, not in a batch at the close. Be willing to throw away themes and answers that the evidence no longer supports, and to commit to the sharper versions as they emerge. The engagements that produce durable recommendations are engagements in which synthesis has been a continuous activity. The ones that struggle are engagements in which analysis was done thoroughly but synthesis was deferred.

Root cause analysis is the second method in the Synthesis section of the toolkit. Where Chapter 51 set out how to turn mess into a structured answer, this chapter sets out how to dig past the visible symptoms of a problem to the underlying causes that, if addressed, would actually move the answer.

The discipline matters because most recommendations that fail to move the business fail because they were addressing symptoms rather than causes. A sales motion is missing its number; the recommendation is to add more salespeople. The number does not move. The actual cause was ICP drift, and adding salespeople has made the unit economics worse. A managed services margin is declining; the recommendation is to take cost out. The margin recovers briefly and then declines again, because the cause was a service definition that has been negotiated away across customer exceptions. The cost programme has bought time without addressing the cause.

This chapter sets out two formal methods that recur in our work — the five whys and the fishbone — and the operator’s instinct that experienced Salamander Associates bring to the same problem.

The five whys.

The five whys method asks “why” repeatedly against a problem until the answer reaches something that can be acted on, rather than something that needs to be explained further. In practice, the number is rarely literally five — sometimes three is sufficient, sometimes seven is required — but the discipline is the same.

A worked example. Why has growth slowed? Because pipeline conversion is lower than it was. Why is conversion lower? Because the qualified pipeline is increasingly outside our ICP. Why is the pipeline outside the ICP? Because marketing has been generating volume without filtering for fit. Why is marketing generating volume without filtering? Because the CMO’s targets are set on lead volume rather than lead quality. Now there is something to address: the marketing target system, not the sales team.

Each “why” in this chain moves the answer one level deeper. The discipline is to keep asking until the answer points to a structural lever rather than a downstream symptom. Stopping too early produces recommendations that treat symptoms; going too deep produces recommendations that are theoretically correct but operationally unactionable.

The fishbone.

The fishbone (or Ishikawa) diagram is the visual companion to the five whys. It maps the potential causes of a problem across categories — people, process, systems, suppliers, environment, measurement — and surfaces causes the team might have missed if working only through a single line of why-questioning.

The discipline in fishbone work is to populate every branch deliberately, not just the branches the team already has hypotheses about. An assessment of a pipeline conversion problem that only looks at the sales team is half-blind; the same problem mapped across people (sales team capability), process (qualification motion), systems (CRM hygiene), data (lead quality), and measurement (target system) usually surfaces causes a single-thread analysis would have missed.

The operator’s instinct.

Formal methods get you most of the way. The senior Salamander Associate’s contribution is often the pattern recognition that comes from having seen the problem before. An ex-CFO walking into a finance function with a forecasting problem usually has a hypothesis about whether the cause is upstream data discipline, mid-stream forecasting methodology, or downstream operating cadence within the first hour. The hypothesis is not infallible, but it is informed, and it shortens the time to a useful root cause materially.

Operator instinct does not replace the formal methods; it complements them. A useful working rhythm is to form the instinct hypothesis first, then test it through the structured method, then update both as the evidence arrives. The Associates who combine the two get to root cause faster than those who use either alone.

Knowing when you have reached the cause.

A few internal checks help distinguish a genuine root cause from another symptom in disguise. Does addressing the candidate cause produce a change that would propagate through the visible symptoms, or does it only address one of them? Does the candidate cause explain the pattern across cohorts, segments, or time periods, or does it only explain the aggregate? If the candidate cause were removed, would the symptoms not recur, or would a similar pattern simply reappear in a different form?

When the answers to these are “propagate,” “explain the pattern,” and “not recur,” the team is at or near root cause. When the answers are “only addresses one symptom,” “only explains the aggregate,” and “would recur,” the team has further to go.

Multi-cause situations.

Most consequential business problems do not have a single root cause; they have a small number of causes operating together. The discipline is not to insist on a single cause but to identify the few causes that, addressed together, would move the problem.

In practice, this means listing the candidate causes from both formal and instinctive analysis, assessing each against the checks above, and converging on the set that the recommendation has to address. Recommendations that name two or three connected causes tend to land better than recommendations that name either one (usually insufficient) or twelve (usually overwhelming and unactionable).

Common failure modes.

A few patterns recur.

Stopping at the first plausible cause. The team has found something that explains the symptom, has not asked whether it explains all the symptoms, and has built the recommendation on that single candidate. The fix is the check list above.

Going so deep that the cause becomes unactionable. The team has traced the problem back to organisational culture, founder mindset, or sector dynamics — true, perhaps, but not something the engagement can move. The fix is to stop at the deepest cause the company can address.

Mistaking pattern for cause. Two things vary together in the data; the team has assumed one causes the other. The fix is the question of whether the candidate cause would propagate the way a real cause would.

Treating cause analysis as one-off. The team has done the analysis at the start of the engagement and has not updated it as more evidence has arrived. The fix is to revisit the cause map at the synthesis stage of the engagement and to be willing to revise it.

For you as a Senior Associate, the practical implication is to ask, for any recommendation you are building, what cause it is addressing and how you know that is the cause rather than a symptom. The discipline of staying at cause rather than symptom is one of the more reliable signals that an engagement is producing durable recommendations rather than ones that will need to be redone in eighteen months.

Scenario thinking is the third method in the Synthesis section of the toolkit. Where root cause analysis (Chapter 52) tells you what is producing the problem, scenario thinking tells you how the answer behaves across the range of futures the company could be heading into. A recommendation that works only in the future the leadership team currently assumes is a fragile recommendation. A recommendation that holds across a range of plausible futures is a durable one. And a recommendation that fails in a particular future identifies the exposure that needs to be addressed.

This chapter sets out how Salamander Associates use scenario thinking, the standard cases that recur, and the discipline of using scenarios to drive decisions rather than to add length to a deck.

The three standard cases.

Most scenario work in Salamander engagements runs against three cases.

The base case. The future the leadership team and the recommendation are planning around. Best estimate of how the business and the environment will evolve, with the assumptions stated honestly.

The upside case. A plausible future in which things go better than the base case — the market grows faster, the new product lands better, the competitive environment eases. Not the most optimistic case anyone could construct; the most optimistic case that the team would actually be willing to defend.

The downside case. A plausible future in which things go worse — the market slows, customer churn accelerates, a competitor moves on price, a key hire takes longer to ramp. Again, not the worst case anyone could construct; the worst case that is plausible enough to plan against.

The discipline is to define each case in terms of the few variables that materially move the answer — the load-bearing assumptions from Chapter 45 — rather than as a vague mood. “Upside” should be a set of specific assumption values, not a feeling that things are going to go well.

When to build scenarios.

Not every recommendation requires scenario analysis. The cases where it earns its keep are recommendations that involve significant commitment of capital, time, or organisational attention, and where the underlying environment is genuinely uncertain. A pricing change that can be reversed in a quarter does not need three cases. A managed services migration that takes eighteen months and reshapes the cost base does.

A useful heuristic: build scenarios where the cost of being wrong is large, the path back is long, and the variables driving the outcome are outside the company’s complete control. Skip scenarios where the decision is small, reversible, or driven by variables the company can directly influence.

When not to build scenarios.

Scenarios become decoration in several situations. When the cases are too close together — the upside, base, and downside differ by margins of error rather than by structurally different futures. When the cases all produce the same recommendation — the analysis has been done but it does not change the decision, which means the analysis was not necessary. When the cases are constructed from optimistic, neutral, and pessimistic moods rather than from concrete assumption variations — the cases feel rigorous but are not.

The discipline is to ask, before building scenarios, whether the recommendation would actually change between the cases. If yes, the scenarios are doing work. If no, the work to construct them is decoration and the time is better spent elsewhere.

How to use scenarios in a recommendation.

A useful Salamander recommendation that has been stress-tested with scenarios does three things explicitly.

It states the base case the recommendation is built against, with the load-bearing assumptions named. It shows how the recommendation behaves in the upside and downside cases. And it identifies the specific exposures — the assumption failures that would break the recommendation, and the early-warning signals that would flag those failures arriving.

The third element is the one most often skipped. A recommendation that says “this works in the base case but is exposed to a slowdown in the segment X cohort, which we would flag through cohort retention dropping below 92 per cent” is more honest, more useful, and more defensible than one that simply produces three numbers without naming what the team should watch.

The base case trap.

A common failure of scenario work is treating the base case as the truth and the other two as embroidery. Leadership teams default to the base case in operating reviews; the downside case becomes a contingency document nobody opens until the contingency has happened.

The Salamander discipline is to give the downside case operational weight. What would have to be true for us to act on the downside scenario? What metrics would tell us we are tracking toward it? What decisions would we make differently if we believed it? Asking these questions during the recommendation, rather than after, is the difference between scenarios that produce resilience and scenarios that produce paperwork.

Common failure modes.

A few patterns recur.

Scenarios that do not change the recommendation. The work has been done; the answer is the same in all three cases. The exercise has consumed effort without producing decision support.

Scenarios that are too close. The cases differ by margins that are within the range of forecast noise. The team has constructed three numbers without three meaningfully different futures.

Scenarios that are moods, not assumptions. The cases are labelled optimistic, neutral, and pessimistic without specifying what is actually different across them. Useless when the leadership team tries to act on them.

The downside case that nobody owns. The scenario was built, the exposure was named, and no one has been assigned the early-warning monitoring or the contingent decision. The work was treated as analysis rather than as planning.

Scenarios that should not have been built. The recommendation is small, reversible, and not subject to significant external uncertainty. The team produced scenarios because the engagement template called for them. Time was lost, recommendation not improved.

For you as a Senior Associate, the practical implication is to use scenario thinking deliberately — where it earns its keep, with cases defined as specific assumption variations rather than moods, and with the downside case operationalised through early-warning signals and contingent decisions. Scenarios that pass these tests are one of the more useful contributions Salamander brings to a leadership team. Scenarios that fail them are paperwork.

The final method in the Synthesis section is the discipline that completes the work — the discipline of pushing every piece of analysis through three questions. What does the analysis show? So what does it mean? Now what should the company do about it? Analysis that stops at “what” is description. Analysis that gets to “so what” is interpretation. Analysis that gets to “now what” is decision support — and decision support is what Salamander engagements are signed up to produce.

This chapter sets out how Salamander Associates use the framework, where it tends to break down, and why the third question is the one that turns advisory work into engagement value.

The three questions.

What. The finding itself, stated clearly. Margin has declined by three points over the last six quarters. Pipeline coverage has slipped from 3.2x to 2.1x. Customer success has been producing more churn than sales is replacing for the last two cohorts. Each of these is a “what” — a statement of what the data is showing, said cleanly enough that the reader does not have to do any work to understand it.

So what. The implication of the finding for the business. If margin has declined by three points, what does that say about the operating model, the competitive environment, the company’s pricing, the cost-to-serve? “So what” pushes the finding toward its meaning. A finding without a “so what” is information; a finding with a “so what” is intelligence.

Now what. The action the finding implies. Given what is true and what it means, what should the company actually do? This is the step that turns analysis into recommendation. It involves the trade-offs the themes themselves cannot make (Chapter 51), the cause-based thinking from Chapter 52, the scenario discipline from Chapter 53. “Now what” is where the judgement of the Senior Associate visibly does its work.

The discipline of completing all three.

The most common failure of analytical work is stopping at “what.” The team has produced a thorough description of what the data shows, has organised it cleanly, and has handed it over without converting it into anything the client can act on. The client has received information, not decision support. The engagement has not earned its keep.

The second most common failure is stopping at “so what.” The team has done the interpretation — the finding is described, the meaning is explained, the implications for the business are laid out — but the action that follows has been left implicit, as if it were obvious. It rarely is. The team has stopped at intelligence; the client still has to do the work of converting intelligence into a decision.

A Salamander recommendation should complete all three for every meaningful finding. The exercise is not always easy — “now what” can be uncomfortable, because it commits the team to a specific course of action that can be examined and challenged. But the discomfort is the work. Engagements that do this consistently produce recommendations that move the business. Engagements that do not, do not.

Applying the framework.

In practice, the framework is most useful when applied at the level of each significant finding, not just at the engagement level. Walk through the analytical work the team has produced and ask, for each piece, what it says, what it means, and what the company should do about it. Where any of the three questions cannot be answered, the analysis is incomplete and needs more work — either more analysis to surface the “what,” more interpretation to push to “so what,” or more judgement to converge on “now what.”

A useful internal habit is to write each finding with all three parts on a single line in a working document — “what / so what / now what.” If the three parts cannot fit on a line, the finding is not yet sharp. If the third part cannot be written at all, the team is not yet ready to put the finding in front of a client.

When “now what” is genuinely deferred.

Occasionally the right answer to “now what” is “we do not yet have enough to commit; here is what we need to know before we can.” This is a legitimate answer, and a more honest one than a “now what” the analysis does not yet support. The Salamander discipline is to be explicit about it — to name what is missing and to design the next phase of work to close the gap — rather than to bury the deferral in vague language.

The discipline is also to recognise when a deferred “now what” is a pattern. A team that defers the third question on every finding is a team that has not yet developed the judgement the engagement requires. The judgement is built by committing to “now what” answers as the evidence permits, taking responsibility for them, and updating as the evidence changes — not by indefinitely deferring.

Common failure modes.

A few patterns recur.

Findings without implications. The deck or memo lists what was found and leaves the reader to do the interpretation. The fix is to push every finding through “so what.”

Implications without actions. The team has interpreted the findings and explained their meaning, but the recommendation has not been made. The fix is to push every implication through “now what.”

“Now what” recommendations that do not flow from the analysis. The team has produced the recommendation in a separate workstream and bolted it on to the analysis. The two do not connect. The fix is to develop them in conversation, as Chapter 49 covered for models and the wider synthesis chapter (51) covered for the full argument.

“Now what” that is too general to act on. “The company should improve its operating model” is not a “now what.” “The company should consolidate its delivery functions into a single managed services unit over the next twelve months, starting with a phase-one design in the next quarter” is. The fix is the discipline of specificity — every “now what” should name what, who, when, and how it will be measured.

For you as a Senior Associate, the practical implication is to ask the three questions of every piece of analytical work you produce, every working memo, every slide, every conversation with the client. The framework is simple to state and surprisingly hard to maintain under engagement pressure, because completing all three commits the team to specific, defensible answers. That commitment is the work. An Associate who is consistently producing “what / so what / now what” findings is producing the kind of work Salamander signs up to deliver.

Communicating top-down is the first method in the Communication and storytelling section of the toolkit. The principle is one of the most important pieces of craft a Senior Associate brings to a Salamander engagement: structure every communication so that the answer comes first, the supporting points come next, and the underlying evidence comes last. Operating people often communicate bottom-up — building toward the answer through context, data, and analysis. Top-down communication inverts that order, and it is the form executives, boards, and sponsors actually want.

This chapter sets out two methods that do most of the work — the Pyramid Principle and SCQA — and the discipline of using them consistently across written and verbal communication.

The Pyramid Principle.

The Pyramid Principle, originally formulated by Barbara Minto, structures a communication as a pyramid. The top of the pyramid is the answer — the conclusion or recommendation the communication is delivering. The middle of the pyramid is the small number of supporting points that, taken together, would lead a reader to the answer. The base of the pyramid is the evidence — data, analysis, examples — that supports each of the supporting points.

The principle that organises the structure is grouping and ordering. Each level of the pyramid groups the elements beneath it according to a logical principle (cause and effect, time order, structure, comparison) and orders them so the reader can follow the argument without getting lost. A pyramid that satisfies this discipline is one in which the reader can read the top sentence, accept it, and stop — or read down through the supporting points, accept them, and stop — or read all the way to the evidence if they want to challenge any point. The communication scales with the reader’s interest and available time.

SCQA.

SCQA — Situation, Complication, Question, Answer — is the companion structure that opens the pyramid. The four elements set up why the communication exists.

Situation. What is currently true. The neutral starting point the reader already understands.

Complication. What has changed, or what has emerged, that disturbs the situation. The reason the communication exists at all.

Question. The question the complication raises, which the communication is going to answer.

Answer. The answer itself — the top of the pyramid.

A good SCQA opener gets the reader from a shared starting point to the answer in three or four sentences. The body of the communication then supports the answer through the pyramid structure beneath.

Why this works for executives.

Executives, boards, and sponsors — the readers of most Salamander communications — share a few characteristics. They have limited time. They have many competing inputs. They tend to be sceptical by default. And they trust the writer more when the writer gets to the point.

A bottom-up communication asks the executive to wait through context, data, and analysis for the answer to emerge. Most executives stop reading before they reach it, or skim ahead, or interrupt with questions the document was going to answer if they had let it. A top-down communication respects the executive’s time and intelligence — it gives them the answer first and lets them dig into the support if they want to challenge it. That respect is one of the reasons clients describe Salamander communications as easier to read than they expected.

Applying the discipline.

The Pyramid Principle and SCQA apply across most forms of Salamander communication.

Written memos. The first sentence states the answer. The next few paragraphs walk through the supporting points. The longer body presents the evidence. A reader who reads only the first page should be able to act on the recommendation.

Decks. The cover slide carries the answer. The executive summary contains the answer plus the three or four supporting points. The body of the deck supports each point with the evidence. The principle is the same across the levels.

Conversations and presentations. The opening sentence states the answer. The middle of the conversation walks through the supporting points. The evidence is held for questions and challenges, not delivered in sequence as if the listener has to earn the answer.

Operating reviews and board meetings. The first words out of the function head’s mouth should be the headline answer for the period — performance against plan, what changed, what to do. The supporting detail follows. Operating reviews that start with twenty minutes of context before the answer is reached have lost the discipline.

Common failure modes.

A few patterns recur.

The answer buried in the body. The communication is technically thorough but the reader has to find the answer themselves. The fix is to put the answer at the top — in writing, in speech, on the first slide, in the first sentence.

The pyramid that does not group. The supporting points are a list rather than a structured argument. They do not connect to each other or to the answer in a way the reader can follow. The fix is the structuring discipline from Chapter 43 — the supporting points have to be MECE against the question they answer.

SCQA that is not actually SCQA. The opening goes situation, situation, situation, situation, with the complication and question never surfaced and the answer left to the end. The reader has read three paragraphs without yet learning what the communication is about.

The pyramid that does not survive translation between media. The memo was written top-down; the deck was built bottom-up; the verbal briefing reverted to chronological. The discipline is to keep the structure consistent across the forms the communication takes.

For you as a Senior Associate, the practical implication is to apply the discipline to every piece of communication you produce, regardless of length or formality. The single-line update at the end of a working session. The memo at the close of a phase. The slide that goes into a board pack. The verbal briefing to the Associate leading the engagement. Every piece of communication should start with the answer, support the answer with a small number of points, and rest the points on evidence. The discipline is simple to state and demanding to maintain. The Associates who maintain it are the ones whose work travels best.

Storylining is the work of building the argument of a communication before building the pages. Where Chapter 55 set out the principles that govern the structure (answer first, supporting points beneath, evidence at the base), this chapter sets out the practical method of producing that structure. The Salamander discipline is to converge on the storyline before any meaningful effort goes into the pages, because pages built on a broken storyline have to be rebuilt, and the rebuild costs more than getting the storyline right in the first place.

This chapter sets out what storylining is, how we do it, and why the headline-level working method is one of the most useful pieces of craft a Senior Associate can develop.

What storylining is.

A storyline is the argument the communication is making, expressed at the level of headlines. Each headline is the answer for a section, not a topic for a section. The complete storyline read through end to end is a coherent argument that takes the reader from the situation, through the complication and the question, to the answer and the supporting reasoning.

A useful test for a storyline is to ask whether a reader who saw only the headlines — the slide titles in a deck, the section headings in a memo — would understand the argument. If yes, the storyline is doing its work. If no, the document is relying on the body content to do work the headlines should already be doing, and the storyline is broken.

The headline-level working method.

Most Salamander communications are built in three phases.

Phase one: the storyline. Working at the level of headlines only. The Associate writes the answer headline, the supporting-point headlines, and the evidence headlines as a sequence of sentences. No pages, no slides, no charts. The storyline can fit on a single page even for a substantial deck.

Phase two: the page sketches. With the storyline agreed, each headline gets sketched — what evidence appears on the page, what visual carries the point, what the page actually shows. Still working in working-document form rather than polished pages.

Phase three: the build. The pages are produced, the charts and tables are drawn properly, the text is polished. By this point, the argument has been agreed at the storyline level and the pages have been sketched. The build is execution, not design.

The phases are not always strictly sequential — engagements often iterate across them — but the discipline is to do the storyline work before the page work, and to revise the storyline rather than the pages when the argument changes.

Horizontal and vertical logic.

A well-storylined deck or memo satisfies two kinds of logic.

Horizontal logic is the logic across the headlines. Read in sequence, do the headlines tell a complete and coherent argument? Does each headline follow from the previous one? Do they collectively support the answer at the top of the pyramid?

Vertical logic is the logic within each page. Does the content on the page support the headline it sits under? Is the chart or table the right one for the point? Does the supporting text add to the headline rather than restating it?

Most communication problems live in the horizontal logic. Pages that individually make sense but do not connect to a coherent argument are the most common failure mode of consulting decks, and the one storylining is most directly designed to prevent.

Building the storyline iteratively.

The storyline is rarely right on the first attempt. The first version sketches the argument as the team currently understands it. The second version refines it after the next round of analysis. The third version, often produced in a working session between the Associate carrying the substantive work and the Associate leading the engagement, sharpens the headlines until each is a clear answer rather than a topic. By the time pages are being built, the storyline has typically been through three or four iterations and is in a form that can support a final document.

The discipline is to do this iteration explicitly, rather than to do it implicitly inside page production. Implicit iteration is what produces decks that have to be rebuilt the week before delivery.

When the storyline is broken.

A few signals indicate that a storyline is not yet ready to support pages.

Headlines that are topics rather than answers. “Sales performance” is a topic. “Sales performance has declined by 12 per cent because pipeline conversion is dropping in the mid-market segment” is an answer. The discipline is to replace topic headlines with answer headlines before pages are built.

Headlines that do not lead to each other. The deck jumps from headline three to headline four without a logical connection. The reader has to do the work of bridging the gap. The fix is to either reorder the headlines or to add the missing link.

Headlines that contradict the answer at the top. The supporting structure does not actually support the conclusion. This usually surfaces when the team has built the conclusion in one workstream and the supporting analysis in another, and they have not yet been reconciled.

A storyline that needs more than one read to understand. The argument is there, but it is buried. The fix is to simplify the structure until the argument is visible on first read.

Common failure modes.

A few patterns recur.

Building pages before the storyline. The team has produced sixty slides of high-quality analysis with no coherent argument across them. The slides are good; the deck is not.

The storyline that exists only in one person’s head. The Associate leading the engagement knows the argument; the team is building pages based on partial briefs. The pages do not connect, because the connecting argument was never made explicit.

The storyline that the analysis no longer supports. The team committed to a storyline early and the evidence has moved; the storyline has not been updated. The pages then have to be retrofitted to evidence that does not fit them.

The storyline that survives review by accident. The Senior Associate reviewing the work was unable to follow the argument but did not say so, and the deck went out without anyone having confirmed it actually held together. The fix is the discipline of pushing back, set out in the integrity value (Chapter 7) and in the constructive challenge work covered in Chapter 61.

For you as a Senior Associate, the practical implication is to storyline before you build, to converge on the headline-level argument before any significant page production, and to be willing to rework the storyline when the argument cannot yet be told cleanly. The Associates whose decks and memos consistently land well are the ones who put most of their effort into the storyline. The pages are easier when the storyline is right.

Writing for executives is the craft of producing prose that lands with people who have little time, high context, and limited tolerance for prose that wastes either. Where Chapter 55 set out the structural principles and Chapter 56 set out the storyline discipline, this chapter sets out the writing itself — sentences, paragraphs, memos — and the rhythm of an executive write-up that gets read and acted on.

The discipline of brevity.

The most common failure of writing for executives is using too many words. A first draft is almost always longer than it needs to be. The second draft cuts the redundant phrasing; the third draft cuts the hedges, qualifications, and throat-clearing that delays the point. A useful internal target is that the final version is half the length of the first draft. If it is not, the writing has not been edited hard enough.

Brevity is not the same as terseness. A short paragraph that gets to the point and supports it cleanly is more useful than either a long paragraph that buries the point or a sentence fragment that omits the support. The discipline is to use the words the argument requires and no more.

Sentence structure.

Executive writing favours short, declarative sentences. The subject is at the beginning, the verb does the work, the object closes it out. Sentences that pile clauses on each other — relative clauses, parenthetical asides, qualifications stacked three deep — slow the reader down and often hide the actual claim.

A useful internal habit is to read each sentence aloud and ask whether the reader could repeat the point after one pass. If not, the sentence is too complex and should be broken into two.

Paragraph structure.

Each paragraph carries one point. The first sentence of the paragraph states the point — the topic sentence, in the Pyramid Principle sense. The remaining sentences in the paragraph support it. The next paragraph starts a new point, with its own topic sentence.

A reader who reads only the first sentence of each paragraph should be able to follow the argument. This is the prose equivalent of the headline-level storyline test from Chapter 56. If the reader cannot follow the argument from topic sentences alone, the paragraphs are not pulling their weight.

The memo structure.

A standard Salamander memo runs to a few pages and follows a consistent rhythm.

The opening. SCQA, compressed to three or four sentences. The situation, what has changed, the question the memo answers, and the answer itself. By the end of the first paragraph, the reader knows the recommendation.

The supporting argument. Three or four paragraphs, each carrying one supporting point. Each topic sentence is itself a complete claim. The supporting sentences provide the evidence.

The implications. A short section on what the answer implies for decisions, timing, and next steps. This is the “now what” from Chapter 54, applied to the memo as a whole.

The appendix or supporting detail. Anything the reader might want to verify but does not need to read in order to act on the recommendation. This is where the evidence base sits.

A memo that follows this rhythm typically lands in three to five pages of body, with the executive summary on the first page able to stand alone for a reader who chooses not to read further.

The executive summary.

The executive summary is the single most important paragraph in a memo or deck, because it is the one most likely to be read in isolation. It should contain the answer in a sentence, the supporting points in a few sentences, and the principal implication for the reader. It should not contain context the reader does not need, qualifications that hedge the answer, or signposts to the body of the document. It should be writable as a standalone piece of communication.

A useful test for the executive summary: would a reader who only read this paragraph be able to act on the recommendation? If yes, the summary is doing its work. If no, it has not yet been written tightly enough.

The language we use.

Salamander prose tends toward operator language — concrete, specific, calibrated to the moment. We name the company, the function, the metric, the decision. We avoid abstractions that could apply to any business — the kind of consulting prose that produces sentences true of every company and useful to none. We avoid hedge words that delay commitment — “we believe,” “it appears,” “potentially” — except where they reflect honest uncertainty, in which case they are appropriate and the calibrated-confidence vocabulary from Chapter 46 applies.

We also avoid jargon, both the consulting kind (“synergies,” “low-hanging fruit,” “deep dive”) and the technology kind (“platform play,” “category creation”) when plain language carries the meaning more cleanly.

Common failure modes.

A few patterns recur.

Throat-clearing openings. The first paragraph reviews the context, restates the brief, and signals what is about to come, before any meaningful content has appeared. The fix is to delete the throat-clearing and start with the answer.

Buried recommendations. The recommendation is somewhere in the middle, prefaced by extensive analysis. The reader has stopped before reaching it. The fix is the SCQA structure, applied honestly.

Hedge stacking. The recommendation is qualified four ways — “we believe it may be appropriate to consider potentially exploring” — because the writer was uncomfortable committing. The fix is the calibrated-confidence vocabulary and a deliberate decision about what level of confidence the underlying evidence supports.

Paragraphs that carry multiple points. Each paragraph is doing two or three things at once; the reader cannot follow the argument because each paragraph is its own mini-thicket. The fix is one point per paragraph, with a topic sentence the reader can extract.

Consulting jargon. The memo reads like a generic strategy deck, with phrases that signal effort more than meaning. The fix is to read every sentence aloud and replace any phrase the writer would not use in conversation with a CEO.

For you as a Senior Associate, the practical implication is to write less than you think the recommendation requires, to cut harder than feels comfortable, and to be willing to put the first sentence of each paragraph through the topic-sentence test. The Senior Associates whose writing lands are the ones whose memos do more with less. The pages are not the work; the argument is, and shorter prose makes the argument harder to ignore.

Visuals on a Salamander engagement exist to drive decisions. A chart, a table, a framework slide, a diagram — each is a tool for getting a specific insight from the analysis into the reader’s head faster than prose would. Used well, visuals accelerate communication and sharpen recommendations. Used poorly, they add length without adding meaning, or worse, they mislead the reader about what the data actually shows.

This chapter sets out how Salamander Associates choose visuals, the question every visual should answer, and the principle of the minimum viable visual.

The question every visual should answer.

For each visual produced on an engagement, the test is whether it answers the question raised by the slide or section it sits in — the headline question. If the slide title says “margin has declined because mix has shifted toward lower-margin segments,” the visual on the slide should show, at a glance, the margin decline and the mix shift. A visual that requires explanation to convey what the headline already states is not doing its work. A visual that conveys information the headline does not name is producing a different argument from the one the slide is making.

This is the vertical-logic test from Chapter 56, applied to visuals.

When to use a chart.

Charts are the right answer when the point is a relationship, a trend, or a comparison that the data can show more cleanly than words can. Revenue over time. Margin by segment. Conversion rates across cohorts. Cost-to-serve by product line. The relationships in the underlying data become visible at a glance.

The standard chart types each have natural homes. Line charts for trends over time. Bar charts for comparisons across categories. Stacked bars or columns for mix at a point in time, or mix shift over time. Waterfall charts for the bridge between two numbers. Scatter plots for the relationship between two variables. Each has a natural use; choosing the right one is a matter of matching the chart to the relationship the data actually contains.

When to use a table.

Tables are the right answer when the point is the actual numbers, particularly when those numbers will be referred to repeatedly. The financial summary that anchors a board pack. The unit economics by customer segment. The scenario comparison across base, upside, and downside cases. A reader who needs to look up a specific number will look it up faster in a table than in a chart.

The Salamander discipline in tables is to design for readability. Heads at the top, totals where they make sense, row and column ordering that supports the argument. Tables thrown together as if the structure does not matter are tables that the reader has to do work to use.

When to use a framework slide.

A framework slide is the right answer when the point is a structure — the components of an argument, the categories of an analysis, the steps in a process. The framework should always carry a headline that states the substantive point the framework is making, not just labelling the framework type. “Our recommendation across the three phases of the managed services migration” is a substantive headline; “Build, Sell, Deliver framework” is a label.

Framework slides are also the type most often misused. A framework that exists only to show that the team has organised its thinking, without adding insight to the underlying point, is decoration. The discipline is to ask whether the framework changes what the reader understands. If yes, keep it. If no, replace it with the substantive content.

When the right answer is no visual.

Some points are best made in prose. A judgement call that depends on context. A recommendation whose force comes from the specificity of its language. An implication that does not lend itself to either a chart or a table.

The Salamander discipline is to be willing to put a slide on the wall with a headline and no exhibit — sometimes a sentence in large type — when that is what the point requires. Slides that produce a visual because slides are supposed to have visuals are slides where the visual is decoration.

The minimum viable visual.

The principle that organises the choice between visual types is the minimum viable visual: the simplest visual that conveys the point cleanly. Most consulting visuals fail by being too complex — too many data series, too many categories, too many annotations, too many colours.

The discipline is to start from the headline of the slide, design the simplest possible visual that supports it, and resist adding anything that does not advance the argument. A chart with two lines and a clear gap between them often makes the point more sharply than a chart with eight lines that requires the reader to find the relevant two.

Common failure modes.

A few patterns recur.

Visuals that do not match the headline. The slide title says one thing; the visual says something adjacent. The reader has to do the bridging work. The fix is the vertical-logic discipline.

Charts that the headline already states more clearly. The visual is decoration. The fix is to either remove it or to redesign it to show something the headline does not already say.

Tables passed off as charts. The data is in a chart format but is so dense that it is effectively a table with axes. The fix is to either convert to a clean table or to filter the chart down to the data that actually carries the point.

Over-annotated visuals. The chart carries labels, footnotes, callouts, and a legend that takes longer to parse than the underlying point requires. The fix is to remove every annotation that does not directly support the argument.

Frameworks without headlines. The slide says “Our framework” or “Strategic options,” without naming the substantive point. The reader cannot extract the argument from the framework alone. The fix is to write the substantive headline first and decide whether the framework supports it.

For you as a Senior Associate, the practical implication is to ask, for every visual you produce, three questions. Does this visual answer the headline of the slide it sits on? Is this the simplest visual that would convey the point? Could a reader who saw only this visual extract the argument? The Associates whose decks land cleanly are the ones whose visuals do work — each one earning its place, none of them decoration.

Interviews are one of the primary evidence sources on a Salamander engagement. The data tells us what the business looks like; conversations with leadership, team members, customers, and partners tell us how it actually runs, what is true that the data has not captured, and what the people inside the company believe is happening — which is often different from what the data shows.

Interview craft is the discipline of conducting those conversations well. This chapter sets out how Salamander Associates approach interviews, the techniques that recur, and the disciplines of listening that turn an interview from a data-gathering exercise into a source of genuine insight.

What we are looking for.

A good interview produces three kinds of information. The factual content the interviewee is willing to share — the data points, the events, the decisions, the names. The interpretation the interviewee places on that content — why things happened, what they mean, who is responsible. And the information the interviewee did not intend to share but that emerges from how they answer — what they hesitate on, what they avoid, what they emphasise.

All three matter. The first is the headline value of the interview. The second is where the interviewee’s perspective adds context. The third is often where the most useful insight lives.

Discovery without leading.

The temptation in a Salamander interview is to use the interview to confirm what the team already believes. The discipline is the opposite — to discover what the interviewee actually thinks, before the interviewer’s view has shaped the conversation.

Open questions do the work here. “What is happening with retention in the enterprise segment?” is a discovery question. “Do you think retention in the enterprise segment is dropping because of our pricing?” is a leading question — it has narrowed the conversation to a hypothesis the interviewee can only confirm or deny.

The discipline is to ask the open question first, and to introduce the team’s hypothesis only after the interviewee has had the chance to describe the situation in their own terms. The interviewee’s framing is often different from the team’s, and the difference is part of what the interview is for.

The structure of a useful interview.

Most Salamander interviews follow a similar shape across the conversation.

The opening sets the context, names what the interview is for, and asks the interviewee for their preferred framing of where to start. Five minutes, no more.

The body of the interview moves from open to specific. The early questions are wide — “tell me about how the function has evolved over the last two years” — and the later questions narrow as the conversation surfaces what is worth digging into. The interviewer is following the conversation rather than running a script, while ensuring the questions the engagement needs answered get asked.

The close summarises what was heard, gives the interviewee a chance to correct, and asks two questions: who else should we speak to, and what would you most want us to understand that we have not asked about. The second question is the one that most often produces material that does not appear elsewhere in the interview.

Listening for what is not said.

The most useful technique in interview work is the discipline of listening for silence, hesitation, and avoidance. An interviewee who answers every question crisply except one — and on that one becomes vague — is signalling something. The signal is not necessarily what the team has hypothesised; it may be discomfort, professional caution, a confidentiality concern. The discipline is to register the signal, hold the moment open without forcing it, and either come back to the topic later in the interview or follow up after.

A second-order version of the same technique is listening for what the interviewee has not mentioned. A function head who describes their function for thirty minutes without ever mentioning a particular team, customer, or topic that the team expected to come up is signalling something. The absence is data.

The follow-up question.

Most interviewers ask the planned question and accept the answer. The Salamander discipline is to follow up — to ask the next question, and the one after that, until the answer has reached something the team did not already know.

“Why” is the most useful follow-up word, though it can sound interrogative if used repeatedly. Alternatives include “what makes you say that,” “what would have to be true for that to change,” “what is the evidence that has shaped that view.” The point is the same: get past the first answer into the second-level reasoning that produced it.

Different interviewee types.

The discipline varies slightly across the people we interview.

Client leadership. The CEO, CFO, function heads. The interviewer should arrive having read the data the leadership team has produced and ready to test hypotheses against the leadership team’s framing. The conversation runs at peer level.

Team members one or two layers down. Often the most useful interviews on an engagement, because the people doing the work see things leadership has stopped seeing. The interviewer should give them room to describe what they actually see, without filtering through how leadership has described it.

Customers. The most useful corrective to the company’s internal view. The interviewer should ask about the customer’s experience, the alternatives they considered, what would make them stay or leave. Customer interviews are also the most likely to produce information leadership genuinely did not know.

Partners and channel. Useful for the view of the company from outside the direct relationship. Channel partners often have a sharper read on competitive dynamics, customer behaviour, and the company’s positioning than the company itself does.

Common failure modes.

A few patterns recur.

Leading questions throughout. The interviewer has spent the hour confirming what the team already believed. The interview has not produced anything new.

No follow-up. The interviewer accepts every first answer at face value. The interview surfaces the headline content and nothing beneath it.

The interviewer talking. The interviewer has spent more than thirty per cent of the conversation explaining the engagement, the framework, or their own view. The interviewee has answered briefly because there was no room to say more.

No closing question. The “what would you most want us to understand” question has not been asked. The most useful material the interviewee was carrying has not surfaced.

The interview that is treated as a single source. The team has built a finding on one interview, without triangulating it against a second. The single-source confidence pattern from Chapter 46 applies.

For you as a Senior Associate, the practical implication is that interviews are one of the highest-leverage activities on an engagement, and the craft of conducting them well rewards practice. Plan the interview, but follow the conversation. Ask open questions before specific ones. Listen for silence, hesitation, and absence. Follow up until the second-level reasoning is on the table. Close with the two questions that produce material nothing else does. The Senior Associates whose engagements produce sharper assessments are usually the ones who conduct interviews well.

Workshops are the other primary form of client engagement on a Salamander project, alongside interviews. Where an interview is the discipline of producing insight from a single conversation, a workshop is the discipline of producing alignment, decisions, or sharpened recommendations from a group conversation. Done well, a workshop is one of the most efficient ways to move a leadership team from a fragmented set of views to a shared answer; done poorly, it consumes hours of senior client time without producing anything the team did not already know.

This chapter sets out how Salamander Associates design and facilitate workshops, the disciplines that produce decisions rather than discussion, and the common ways this kind of work goes wrong.

When to run a workshop, and when not to.

The first decision is whether a workshop is the right format at all. Workshops earn their place when a question requires multiple stakeholders to be in the room together, when the answer depends on alignment that cannot be assembled through one-to-one conversations, or when a decision needs to be made collectively and committed to in the meeting itself.

Workshops do not earn their place when a question could be answered through a few interviews, when the input could be gathered asynchronously, or when the workshop is being scheduled because someone thinks the team should “have a workshop” rather than because the question requires one. A workshop that does not need to be a workshop is an expensive way to do work that could have been done faster.

Workshop design.

Most of the value of a workshop is determined before the workshop begins. The discipline is to do the design work first — what the workshop is trying to produce, who needs to be in the room, what materials they need beforehand, how the time will be used — rather than to assemble the workshop and rely on facilitation skill to make it work.

A useful design checklist runs across a few elements.

The output. What does the workshop need to produce — a decision, an alignment, a sharpened assessment, a workplan? If the output cannot be named in a sentence, the workshop has not been designed.

The pre-read. What participants need to have read and processed before the workshop, so the time in the room is not spent briefing people on context. Most workshops that disappoint do so because participants were under-briefed in advance.

The participant list. Who needs to be in the room to make the output possible, and who would be better consulted before or after. Workshops that include people whose role does not match the output produce diluted conversation.

The agenda. The flow from opening to output, with each segment named for what it produces, not what it discusses. “Decision: which of the three options we go with” is an agenda item; “Discussion of strategic options” is not.

The materials. The visuals, framings, and exhibits the workshop will work against. Materials prepared in advance, agreed by the engagement team beforehand, and shaped to support the workshop’s output.

The structure of a working session.

A typical Salamander workshop runs against a structure that has the output at its centre.

The opening positions the workshop — what it is for, what success looks like, what is being decided. Five minutes.

The grounding segment establishes the shared starting point — the assessment, the data, the prior decisions the workshop is building on. This should be brief and pre-read; the workshop is not the place to teach.

The body of the workshop is structured around the question the workshop is answering. For a decision workshop, this is the structured comparison of options. For an alignment workshop, this is the surfacing of differences and the work to converge. For an assessment workshop, this is the structured walk through the evidence and what it implies.

The decision point is named explicitly. The workshop is paused, the question is restated, the decision is made, and the commitment is captured.

The close summarises what was decided, what the next steps are, and who is responsible for each. Workshops that end without an explicit close tend to produce work that the participants remember differently a week later.

Facilitation craft.

The Salamander Associate facilitating a workshop is responsible for several things at once. Keeping the conversation on the output. Drawing out participants whose views matter and who have not yet spoken. Containing participants who are taking more time than their input warrants. Surfacing disagreement honestly rather than letting it sit unaddressed. Reading the room for the moment a decision could be made versus the moment more conversation is genuinely required.

A useful internal discipline is to be willing to interrupt the conversation. A workshop that runs as a free-form discussion will reach the end without producing the output. A workshop that the facilitator structures actively — naming what is being discussed, when it is time to move on, when a decision is in front of the group — is one that produces decisions.

Producing decisions, not just discussion.

The single most common failure of workshops is producing discussion without producing decision. The group has talked productively for ninety minutes; the participants leave with a sense that the conversation was useful; nothing was actually decided.

The Salamander discipline is to be willing to put a decision in front of the group explicitly and to hold the room there until the decision is made or until it is honestly named as one that cannot yet be made and therefore requires a different forum. “We have heard both views; the decision in front of us is X; can we commit to that, with these terms, in this meeting?” is a sentence that workshops often need someone to say. The facilitator’s job is to be the person who says it.

Common failure modes.

A few patterns recur.

Workshops without named outputs. The team has scheduled the workshop without being able to say what it is supposed to produce. Discussion happens; output does not.

Under-prepared participants. The pre-read was sent late, was too long, or was never produced. The workshop spends its first hour briefing people who should have been briefed before.

Too many people in the room. The decision the workshop is trying to make needs four people; nine are present. The conversation dilutes; no decision lands.

Facilitation without structure. The conversation is allowed to flow without anyone holding it to the output. Participants leave feeling the time was useful and the engagement team has nothing to show for it.

No close. The workshop ends when time runs out, without an explicit summary of what was decided and what comes next. Different participants leave with different impressions of what was agreed.

For you as a Senior Associate, the practical implication is to invest most of the workshop work in the design phase. A well-designed workshop is easy to facilitate; a poorly-designed one cannot be rescued in the room. Name the output. Pick the participants. Prepare the materials. Build the agenda around what is being produced. The hour in the room is the visible part; the hours of design before it are where the value is actually built.

Constructive challenge is part of what Salamander engagements are signed up to deliver, and it sits inside the integrity value (Chapter 7). We are paid partly to tell clients things they would rather not hear, partly to push back on internal logic that has stopped serving them, and partly to challenge the framing the leadership team has settled into. The challenge is part of the work. The discipline is to do it in a way that strengthens the relationship rather than damaging it.

This chapter sets out what constructive challenge looks like in practice on a Salamander engagement, how the technique differs between challenging clients and challenging colleagues, and the calibration that turns honesty into something useful rather than something corrosive.

Why constructive challenge matters.

Clients hire Salamander partly for the assessment and the recommendation, but also for the willingness to push back on positions the leadership team has held for too long without examination. An advisor that nods through everything is not adding the value the client engaged us for. The integrity value says it directly: we communicate openly, we challenge constructively, we own outcomes. The constructive challenge piece is not optional; it is part of the contract.

There is also a self-interested case. Salamander recommendations are more durable when they have been tested against the strongest objections inside the company. The leadership team that has heard the challenge in the engagement — and either agreed with the underlying point or talked the team out of it — is more committed to the recommendation than a leadership team that has been told what to think. Constructive challenge is one of the ways the engagement earns the leadership team’s ownership of the answer.

The calibration question.

Chapter 7 set out the anecdote that frames the calibration. Salamander once lost a client because someone on the team was too direct — treating the client like a team member to be run rather than a leader to be worked alongside. The lesson is the calibration: the honesty stays the same; the register has to shift.

Constructive challenge respects the difference. The Salamander Associate is not the client’s boss; the client is not the Associate’s report. The challenge is being delivered to someone whose decision it ultimately is, not to someone whose answer the Associate is owed. The register reflects that — direct, but respectful; clear, but inviting rather than demanding; pointed at the substance, not at the person.

How to challenge well.

A few techniques recur in effective constructive challenge.

Lead with the question, not the verdict. “How would you respond if a buyer’s diligence team asked about the segment cohort retention?” is a useful challenge — it invites the leadership team to surface the answer themselves, often arriving at the discomfort the Associate would otherwise have to deliver as a verdict.

Anchor in the evidence. “The cohort data shows retention falling in the segment we are pursuing most heavily. What is producing that, and what is the implication for the strategy?” The challenge sits inside the data; it is not the Associate’s personal view against the leadership team’s. The conversation moves from “the Associate disagrees with us” to “what does the data show, and what does it imply.”

Name the trade-off the recommendation depends on. “This recommendation works if pricing holds; the historical pattern suggests it may not. What would have to be true for the recommendation to survive a price test?” The challenge is structural — it surfaces the load-bearing assumption from Chapter 45 rather than the personal disagreement.

Acknowledge what is already working. The challenge that says only what is wrong feels destructive. The challenge that says “the X piece is working; the Y piece is the one I want to push on” is more easily heard. The first half is not flattery; it is calibration.

Give the client time. A challenge that lands well is often a challenge that the client gets to absorb, push back on, and revisit. The Associate who makes the point and then keeps adding to it leaves no room for the client to do the work the challenge was designed to provoke.

When to challenge, and when to hold back.

Not every disagreement is worth surfacing. The discipline is to ask, for any potential challenge, whether it matters to the recommendation or to the integrity of the engagement. If yes, raise it. If no, let it go.

A useful internal filter is the load-bearing test. Challenges that touch a load-bearing assumption (Chapter 45), a major decision the engagement is signed up to support, or an integrity issue in how the work is being represented should be raised. Challenges that are personal preferences, stylistic differences, or matters of internal client culture that the engagement does not need to change usually should not.

Challenging clients versus challenging colleagues.

The same value applies inside Salamander, but the calibration is different. Chapter 7 set out that the same direct-but-respectful exchange applies between Salamander Associates regardless of who is challenging whom. With clients, the operator-running-the-team register would damage the relationship. With colleagues, the holding-back-because-of-rank register would damage the work.

The Salamander discipline is to read the audience. The substance of the challenge — what is being said — is the same. The register — how it is being delivered — varies between the colleague-to-colleague channel and the Associate-to-client channel.

Common failure modes.

A few patterns recur.

Challenge that is delivered as verdict. The Associate has named the disagreement as if it were a settled conclusion. The client experiences this as confrontation, regardless of how correct the underlying point is. The fix is the question-form delivery.

Challenge that is held back. The Associate has noticed the issue and not surfaced it. The recommendation goes out with a known weakness unexamined. The fix is the load-bearing filter: if the challenge matters, it is part of the work to raise it.

Challenge that turns personal. The conversation has moved from the substance to the people. The relationship is damaged; the underlying point is lost. The fix is to anchor in evidence and to direct the challenge at the framing or the assumption, not at the individual.

Challenge accumulated for the end. The Associate has saved several challenges for a final meeting, where they land as a sequence of objections. The fix is to raise challenges as they arise, so the leadership team can absorb them one at a time rather than all at once.

Challenge inside Salamander filtered through caution. An Associate who sees a problem in a colleague’s work has held back because the colleague is more experienced, more established, or because the engagement is going well. The fix is the same direct-but-respectful exchange Chapter 7 described — Salamander Associates are at the same level regardless of external history, and the discipline of challenge applies in both directions.

For you as a Senior Associate, the practical implication is to challenge when the substance warrants it, deliver the challenge in the register the audience requires, and treat constructive challenge as part of the work rather than an exception to it. The Associates whose clients respect them most are usually the ones who have pushed back honestly and respectfully when the recommendation called for it. The Associates who get into trouble are usually either the ones who push back too aggressively, or the ones who do not push back at all.

Reading stakeholders and the boardroom is the discipline of understanding the political landscape inside a client, identifying who actually makes decisions, and avoiding the landmines that can derail a technically correct recommendation. As Chapter 19 noted, work that is technically right but politically misplaced is one of the more common failure modes in sponsor-led engagements; the same is true in most other engagements where multiple stakeholders are involved.

This chapter sets out how Salamander Associates read the stakeholder map of a client, the dynamics that recur in B2B technology companies, and the practical disciplines that turn political awareness into engagement effectiveness — without turning Salamander into a political operator inside the client.

The basic stakeholder map.

On most engagements, the people who matter sit across a few categories.

The decision-maker. The single person whose decision the engagement is signed up to support. Often the CEO, sometimes the CFO, sometimes the head of the function the engagement is most directly working with. Identifying the decision-maker on day one is one of the most consequential reads on an engagement.

The decision-influencers. The two or three people whose views meaningfully shape the decision-maker’s view. Often function heads, sometimes board members, sometimes key investors, sometimes a trusted external advisor the client speaks to outside the formal process.

The decision-blockers. People who have the capacity to slow or veto a decision they disagree with, even if the formal authority is not theirs. Senior individuals whose buy-in is needed for implementation, board members with informal influence, partners whose support matters operationally.

The decision-affected. Everyone who will live with the outcome. Their views matter both because they often have insight the leadership team has lost, and because their willingness to execute determines whether the recommendation lands.

A working stakeholder map of an engagement names each of these and identifies how the engagement is engaging with each. A recommendation that has been sold to the decision-maker but not to the decision-influencers usually does not survive contact with the wider organisation.

Getting to the actual decision-maker.

The formal decision-maker is sometimes the actual one and sometimes not. A CEO who has appointed a head of strategy and consistently defers to them on strategic recommendations has effectively moved the decision. A function head who reports to the CEO but has the board’s confidence on a particular topic may carry more weight on that topic than the CEO does. A private equity sponsor who is formally one step removed from operating decisions may in practice be the person who decides the recommendation has to be reshaped before the leadership team will adopt it.

The Salamander discipline is to read for the actual decision-maker rather than to accept the formal one. The reads that surface this include: who does the CEO refer to before committing? Whose objections will the CEO not override? Who does the leadership team look at when the question is raised? The answers are often visible within the first two or three operating reviews if the Associate is reading the room rather than the data.

The boardroom dynamic.

Board dynamics differ from leadership team dynamics. Boards are usually composed of investors, independents, the CEO, and sometimes one or two other senior executives. The dynamics that matter include the relationships between the CEO and the chair, between the investors and the independents, between any single dominant investor and the others, and between the board and the management team it has hired.

Recommendations that have to land in a boardroom should be read through this lens. A recommendation that makes commercial sense to the CEO but that one of the independent directors has previously argued against will struggle, regardless of its merit. A recommendation that the CFO has not been consulted on, in a board that is reading the company through the CFO’s framing, will face resistance from the CFO’s silence even more than from a stated objection.

Avoiding landmines.

A landmine in a B2B technology engagement is a sensitivity the engagement team has not registered until the conversation has stepped on it. Common varieties: prior failed initiatives that the current recommendation looks like; senior individuals whose history with a topic predisposes them against any move on it; investor positions that the leadership team has been careful not to challenge; cultural or regional sensitivities that an outside team is unlikely to read correctly without help.

The discipline of landmine avoidance is to surface them deliberately in the first few weeks. Useful sources include the client’s most candid leadership team members, the operating partner or sponsor where one exists, and prior engagements with the same client where Salamander has run before. Asking openly — “what should we not say in front of the board,” “what topic does the CEO not want raised yet,” “what previous initiative is this likely to be compared to” — saves the engagement weeks of work that would otherwise be undone by an avoidable misstep.

The line between political awareness and political operating.

Reading stakeholders well and operating politically are different. The Salamander discipline is the first, not the second. We read the political landscape to deliver a technically right recommendation in a way that lands, not to soften the recommendation to please the politics. A recommendation that has been weakened to suit the stakeholder map has failed the integrity value (Chapter 7). A recommendation that has been delivered to the stakeholders in the right order, with the right framing, with the right people consulted in advance, has used political awareness to do the substantive work better.

Common failure modes.

A few patterns recur.

Reading the formal decision-maker as the actual one. The engagement has been calibrated to the wrong person; the actual decision-maker is hearing the recommendation second-hand.

Stakeholder map that has not been updated. The map from week one is still being used in week eight; the situation has moved.

Recommendations sold up but not across. The decision-maker has been brought along; the decision-influencers have not been consulted. The recommendation fails when the influencers raise objections in the meeting where it was supposed to land.

Landmines discovered by stepping on them. The engagement has surfaced a sensitivity in the boardroom rather than in pre-conversations. The cost is paid in the relationship as well as in the work.

Political operating. The Associate has softened the recommendation to suit the politics, rather than using the politics to deliver the unsoftened recommendation in the right way. The integrity value has been compromised.

For you as a Senior Associate, the practical implication is to build a working stakeholder map on every engagement, update it as the engagement moves, and use it to shape how the work lands rather than what it concludes. Read the actual decision-maker, not the formal one. Surface landmines early. Sell across as well as up. Treat political awareness as a craft that supports the work, not as a substitute for it. The Associates whose recommendations land well are usually the ones who have done this read with care.

KPIs and operating metrics are the working currency of every Salamander engagement. The discipline of picking the right ones, knowing what each one is for, and not getting buried under the ones that do not earn their place is one of the more important pieces of craft a Senior Associate develops. Most B2B technology companies we walk into have either too few metrics (and are therefore running on instinct) or too many (and are therefore running on dashboards no one reads).

This chapter sets out how Salamander Associates work with metrics in practice. Chapter 64 covers the related craft of building metric trees, which extends what is described here.

The metric proliferation problem.

Most scaling B2B technology companies accumulate metrics over time. A new function head joins and instruments the metrics they ran in their previous role. A new system is implemented and produces fresh dashboards. A board meeting raises a question that generates a recurring metric to answer it. Three years later, the company has a hundred and twenty metrics, ten of which are looked at in leadership meetings, three of which actually drive decisions, and a long tail that absorbs reporting effort without informing the business.

The Salamander discipline is to pick the small set that matters and to put the rest aside. Most well-run scaling companies operate on between five and fifteen KPIs at the leadership level, with deeper metric trees beneath them for the function heads who need the operational view.

The Three M’s: Model, Metrics, Management.

A useful Salamander framing for sequencing metric work is the Three M’s — Model, Metrics, Management. The principle is that the three have to be addressed in the right order for any of them to produce real discipline.

Model first. The financial model that captures how the business actually works — the drivers, the unit economics, the levers — has to be sound before any meaningful metric design can sit on top of it. A KPI set built against a model that does not reflect the business produces metrics that do not move the right levers.

Metrics second. With the model in place, the metric set follows. The headline KPIs at the leadership level, the metric trees beneath them (Chapter 64), the leading and lagging indicators, the dashboards that surface them. The metric design flows from the model rather than from intuition about what to measure.

Management third. The leadership behaviour, the operating cadence, and the incentive system that runs against the metrics. Without disciplined management of the metrics, the model and the metric set are documentation — they do not change the business. With it, the metric set becomes the operating system of the leadership team.

The sequencing matters. Companies that try to install metrics without fixing the model first produce noise. Companies that install metrics without management discipline produce dashboards. Only the three together produce the operating discipline that scaling companies actually need.

Picking the right small set.

A useful KPI satisfies a few criteria.

It is tied to a decision. The metric exists because there is a decision someone makes against it. A metric without an associated decision is a metric that consumes reporting effort without producing information that matters.

It is measurable consistently. The definition is stable, the data source is reliable, the calculation produces the same answer every period.

It is responsive to action. The leadership team can do something to move it within a relevant time frame. A metric that no decision can influence in the engagement’s time horizon is informational rather than operational.

It is honest about lag. The metric measures something the team can actually see at the cadence claimed. A “weekly NPS” that depends on a quarterly survey is misleading; a weekly metric should be weekly data.

These four criteria do most of the work. Metrics that satisfy them are KPIs in the operational sense. Metrics that satisfy fewer are reporting artefacts.

The Balanced Scorecard lens.

A useful counterweight to KPI sets that have drifted toward a single dimension is the Balanced Scorecard framing, formulated by Kaplan and Norton. The framework measures performance across four perspectives — financial, customer, internal process, and learning and growth — and treats the four together as a more complete view of the business than the financial perspective alone.

The discipline matters because scale-ups in particular tend to over-weight the financial perspective (revenue, margin, cash) and under-weight the operating perspective (customer health, delivery quality, organisational capability). A KPI set that runs only on the financial perspective produces a leadership team that sees the consequences of operating problems after they have already arrived in the numbers. A scorecard balanced across the four perspectives produces a leadership team that catches the operating signals earlier.

Salamander engagements often include rebuilding a leadership team’s metric set against this lens — adding the leading indicators from the customer and operating perspectives, and connecting them to the financial lagging indicators they will eventually move.

A practical calibration. Full Balanced Scorecard implementation — the formal four-perspective scorecard with cascade through the organisation — is overkill for an early-stage company. A Series A or Series B scale-up does not need the apparatus; it needs a small number of well-chosen metrics across each of the four dimensions, with the discipline to look at them together rather than treating them as competing functional reports. The framework’s value at that stage is the lens, not the implementation. As the company scales, the formal scorecard becomes more useful — but the Salamander discipline is to introduce it deliberately at the size and complexity where it earns its place, rather than imposing it as a template.

Lagging and leading indicators.

A lagging indicator measures what has already happened — revenue, margin, customer churn, EBIT. These are the headline metrics every leadership team tracks, and they are the metrics boards and investors use to judge the business.

A leading indicator measures something that predicts the lagging indicator — pipeline coverage, customer health scores, employee engagement, product usage trends. The leading indicator changes before the lagging one does, which is what makes it useful for action.

A working KPI set has both. Lagging indicators alone tell the team what has happened, by which point the action that would have changed it is too late. Leading indicators alone produce activity without confirmation that the activity is producing the intended outcome. The combination — leading indicators that the team operates against, lagging indicators that confirm the leading indicators were worth running on — is the working set.

Output and activity metrics.

A related distinction is between output metrics (what the function is producing) and activity metrics (what the function is doing). Output metrics are usually more meaningful — pipeline generated, deals closed, customer health improved — but activity metrics can be useful as early indicators of output if the relationship between activity and output has been tested.

The discipline is to remember the difference. A sales team that is hitting its activity targets (calls made, meetings held) but missing its output target (revenue closed) has an activity-output disconnect that the activity metrics are not surfacing. The Salamander work is sometimes to install or refresh the output metrics in functions that have drifted into activity-only measurement.

The metric purpose question.

For every metric in an operating review or board pack, the question worth asking is: what is this metric for? The honest answers fall into a few categories.

To monitor health. The metric is reported because the leadership team needs to know whether something has changed. No specific decision is being made against it most of the time; the metric is on standby.

To drive a decision. The metric is reported because the leadership team makes a specific decision based on it — investment level, capacity, hiring, pricing.

To track an initiative. The metric is reported because a specific programme of work is being measured against it — usually time-bound, with the metric retiring when the programme concludes.

To inform an external audience. The metric is reported because a board, investor, or sponsor expects to see it — sometimes useful for the team, sometimes only ceremonial.

A well-organised KPI set has metrics in each of the first three categories with the fourth kept to a minimum. Most companies we walk into have the fourth category bloated and the third category neglected.

Dashboards that work.

A dashboard that works has a few characteristics. It has fewer metrics than the team initially wanted. Each metric on it is tied to a decision, an initiative, or a health monitoring purpose. The metrics are arranged so the reader can see the headline picture at a glance and drill into specifics where the headline raises a question. The dashboard is updated on the cadence the metrics support, not on a frequency that produces stale numbers.

The dashboard that does not work has the opposite characteristics. Too many metrics, none of them clearly tied to decisions, all displayed at uniform weight, updated on a cadence that mixes daily, weekly, monthly, and quarterly numbers without distinguishing between them. The Salamander work in operating-cadence engagements often includes stripping the dashboard back to the working set.

Common failure modes.

A few patterns recur.

Metrics without decisions. The pack is full of numbers, but no decision moves with them. The fix is to retire any metric that cannot answer “what decision does this drive?”

Lagging-only measurement. The team is running on revenue and margin and discovers problems six weeks after they have started. The fix is to install leading indicators for the dynamics the team needs to manage.

Activity measurement without output anchoring. The function is busy hitting its activity targets; the business is not producing the output the activities were supposed to drive. The fix is to anchor the activity metrics in output metrics and to question the activity targets when the output target is missed.

Dashboards that nobody reads. The dashboard exists; the leadership team has stopped opening it. The fix is to redesign the dashboard around the small number of metrics the leadership team actually needs to act on.

For you as a Senior Associate, the practical implication is to read every metric set you encounter against the criteria above. Which metrics are tied to decisions? Which are leading versus lagging? Which are tracking outputs versus activities? Which exist because they always have, with no current purpose? The Senior Associates whose engagements produce useful metric work are the ones who push back on the existing set as much as they add to it.

A metric tree is the structured decomposition of a top-line number into the operational levers that produce it. Where Chapter 63 set out how to choose the right metrics, this chapter sets out how to structure them so that a leadership team can trace performance from the headline result back to the drivers actually moving the business.

Metric trees are one of the more useful pieces of craft in the toolkit, because they connect analysis to execution. A leadership team that can see the tree understands not just what is happening but where to act on it. This chapter sets out how Salamander Associates build trees, how OKRs fit alongside them, how to attach incentives at the right level, and the figure that makes the structure visible.

What a metric tree is.

A metric tree starts with a top-line number — revenue, EBIT, NRR, customer acquisition cost — and decomposes it into the drivers that combine to produce it. Each driver, in turn, decomposes into its own drivers, until the leaves of the tree are the operational levers an individual function head can actually move.

The tree is not a list of metrics. It is a structured argument about how the business works. Read top-down, it tells the leadership team what produces the top-line. Read bottom-up, it tells the function heads how their work contributes to the company’s outcomes. The two readings are the same tree, used by different audiences.

The illustration above is a simplified worked tree for a B2B technology revenue line. A working tree for a specific client engagement would go deeper, with the leaves of the tree being the actions the team can take week to week.

Building a tree from a top-line number.

The discipline of building a tree is to decompose deliberately rather than to list everything that might affect the top-line.

Start with the top-line and ask: what additive or multiplicative components produce it? Revenue equals new revenue plus expansion revenue plus retained revenue. EBIT equals revenue times gross margin minus operating expenses. Each top-line metric has a natural first-level decomposition that the engagement should pick deliberately rather than borrow from a template.

Each first-level driver gets the same treatment. New revenue equals pipeline times conversion times average selling price. Pipeline equals leads times qualification rate. Conversion equals stage-one to stage-two rate times stage-two to closed rate. Each level adds resolution until the leaves are operational.

Two disciplines keep the tree useful. The first is MECE (Chapter 43) — each level’s drivers should be non-overlapping and exhaustive. The second is the actionability test — each leaf should be something a function can act on within the relevant time horizon. A leaf that no one can move is not a leaf; it is a description.

How OKRs fit into a metric tree.

Most B2B technology scale-ups operate on Objectives and Key Results rather than (or alongside) KPIs. The two structures are complementary, not competing, and a leadership team running on OKRs is using a particular form of the same metric tree discipline.

An objective is an outcome statement at the top of the tree — what the team is trying to achieve in the period. Key results are the measurable indicators that prove the objective has been reached. In tree terms, the objective sits at the top, the key results sit beneath as the small set of drivers the team commits to moving, and the operational leaves beneath those are the work that produces the key results.

A useful Salamander frame: KPIs and metric trees describe how the business works; OKRs describe what the leadership team is committing to in the period. The tree is the architecture; the OKRs are the focus inside it. A scale-up that has both — a stable metric tree describing the business and a quarterly set of OKRs naming where the leadership team is putting its effort — has the discipline that runs well at the size most of our clients sit at.

OKRs that are not connected to the underlying tree tend to drift. Quarterly objectives become disconnected from the business model, key results become activity targets rather than outcome measures, and the OKR ritual produces noise rather than focus. The Salamander work in OKR-based companies is often to install or refresh the tree the OKRs are supposed to sit inside.

Aligning metrics, outcomes, and incentives.

A metric tree only changes the business if the people inside the business are paid against the right levels of it. Misaligned incentives are one of the more reliable reasons recommendations fail to land — a leadership team commits to a tree, the team beneath them is incentivised against different metrics, and within two quarters the behaviour has drifted back toward what the incentives reward.

The discipline is to attach incentives at the right level of the tree. Function heads should be incentivised against the metrics they can move — usually the second or third level of the tree, not the top line and not the operational leaves. A CRO incentivised against pipeline alone is incentivised against an activity metric; a CRO incentivised against pipeline and conversion and ASP, weighted against the new revenue line they actually produce, is incentivised against outcomes.

The Salamander work on incentive alignment usually involves three moves. Reading the existing incentive structure — what people are actually paid against. Mapping it against the metric tree — which levels are covered, which are not, and where the misalignment sits. And redesigning the structure so the incentives pull in the same direction as the tree. Companies that complete this work see behavioural change within two quarters; companies that build the tree without addressing the incentive structure usually do not.

This connects to the Three M’s from Chapter 63. Model produces the tree. Metrics populate it. Management runs against it — and the incentive system is one of the principal Management tools. The three together produce the operating discipline; the tree by itself does not.

Common failure modes.

A few patterns recur.

Trees that are too deep. The team has produced six levels of detail when three would have served. The leadership team cannot use the tree because the resolution exceeds the decision-making level.

Trees that are too shallow. The tree stops at the function-head level without going down to the operational levers. Function heads cannot use the tree because the next level of detail — the one they actually need — is not there.

Trees disconnected from the data. The tree exists on paper but the underlying data systems do not produce the metrics the tree requires. The tree is conceptual; the operating reality does not match it.

OKRs without a tree. The quarterly OKR set does not connect to a stable underlying view of the business. Each quarter produces new objectives that have no relationship to the prior quarter or to the business model.

Incentives misaligned with the tree. The work was done on the tree; the incentive structure was not revisited. Within two quarters, behaviour drifts back to whatever the incentives reward.

For you as a Senior Associate, the practical implication is to build the tree, connect it to whatever OKR structure the client operates, and check the incentive system before the engagement closes. A metric tree that is technically sound but commercially disconnected — because the incentives pull against it — is a tree that documents the business rather than running it. The Associates whose metric work actually changes how a company operates are the ones who do the alignment work alongside the tree work.

That completes Part 7. The next part of the guide turns from the analytical and advisory toolkit to the practical mechanics of running a Salamander engagement.

Scoping is the work of turning a client conversation into an engagement that can be delivered. It is also one of the highest-leverage activities on any Salamander engagement, because most engagements that succeed are scoped well at the outset, and most engagements that struggle were scoped poorly. The hour or two spent getting the scoping right at the start usually saves weeks of correction later.

This chapter sets out how Salamander Associates approach scoping, what a good scoping document contains, and the disciplines that turn an initial conversation into a workable shape.

The first conversation.

Most Salamander engagements begin with a conversation between the prospective client (usually the CEO, CFO, or relevant function head) and a Salamander Associate — often someone who is going to lead the engagement, or who is positioned to. The conversation is part discovery and part scoping. The Associate is listening for what the client actually needs, what they think they need, what they have already tried, and what the constraints around the engagement might be.

A useful structure for the first conversation runs across a few questions. What is the situation, in the client’s own framing? What has changed that produced the need for an outside view? What outcome would make the engagement worth doing — what does the client want to be true at the end of it? What is the timeline driving the work — a board cycle, a fundraise, a transaction, an operating problem getting worse? Who else needs to be involved, and what do they think? What has been tried so far, and why has it not yet solved the problem?

The Associate should leave the first conversation with enough to draft an initial scoping document — even if not enough to sign one.

What a good scoping document contains.

A Salamander scoping document includes the elements a serious engagement requires. It is not a one-page note; it is the document that defines what the engagement will deliver, when, and to what standard.

The outcome. The single most important section. What state the business needs to be in at the end of the work, expressed in terms specific enough that both Salamander and the client will know whether the engagement has delivered. As Chapter 12 set out, the emphasis is on the outcome — the deliverables and timelines that follow are scoped to serve it.

The scope of work. What the engagement will cover, with the boundaries named clearly. What is in scope; what is explicitly out. Engagements that drift do so because their scope was set as a topic rather than as a defined piece of work, and the topic admits indefinite expansion.

The deliverables and timeline. A Salamander scoping document does include these clearly. They sit underneath the outcome — what the client will see and when, and how those artefacts contribute to the outcome the engagement is signed up to produce.

The team and the model. Who is leading, who is supporting, what fractional capacity each person is committing to the engagement, what the engagement looks like in days per week or month over the duration. The fractional model from Chapter 15 makes this section more substantive than a traditional consulting scoping document — the client is committing to a senior person’s time at a specific cadence.

The commercial terms. Fees, payment schedule, expenses, success components where they apply. Stated clearly enough that there are no surprises later.

The dependencies and assumptions. What has to be true on the client side for the engagement to work — access to data, leadership time, decisions the client will need to make. Recommendations that depend on assumptions the engagement cannot test are flagged here too.

The early termination terms. How the engagement can be wound down if either side concludes it is not working. Adult provisions in adult engagements.

The Salamander scoping toolkit.

Salamander maintains a set of assessment tools on the intranet dashboard (Chapter 71). The one Senior Associates use most often at the start of an engagement is the online assessment tool, and it does two things from the same inputs.

The first is to generate a report for the client — a structured read of where the company sits against the relevant Salamander archetype, based on the discovery conversation, the data the client has shared, and the responses to the tool’s assessment questions. The report is part of what the client receives in the scoping process; it gives them a snapshot of the assessment the engagement is built around.

The second is to generate the draft scope for the engagement — the working document the Associate then edits and completes. The draft is a starting point, not a finished document. The Associate sharpens the outcome statement, calibrates the scope against the specific client, adjusts the fractional capacity to the engagement’s actual shape, and confirms the dependencies and assumptions. This human-in-the-loop step is where the Associate’s judgement does the substantive work.

Once the scope is complete, the tool generates the quotation, the standard Salamander terms and conditions, and the cover letter from the agreed scope. The commercial output flows from the scope rather than being produced independently — which is part of why getting the scope right matters.

Using the assessment tool from day one of an engagement is part of how Salamander runs efficiently. It also ensures the engagement uses the standard Salamander commercial terms and the consistent format clients have come to expect. Senior Associates should make a point of learning the tool early — the time it saves on every engagement compounds over the first year.

Shaping the engagement.

Scoping is partly definitional and partly design work. The shape of the engagement — how long it runs, who is involved, how it phases, what cadence it operates on — is part of the value the Associate brings to the conversation.

A few shaping decisions recur. Phase versus continuous engagement: is this a discrete piece of work with a clear endpoint, or a fractional role that runs until a permanent successor is recruited? Single-Associate versus pair: does the engagement need one Salamander person or two, and if two, in what configuration? Front-loaded versus continuous cadence: should the engagement go heavy in the first six weeks and lighten, or run at steady intensity across the full duration?

These decisions should be made deliberately at scoping. Engagements that begin without these shaping calls explicit tend to drift into whatever the workload demands week to week, which is usually not the most efficient shape.

Pricing and economics.

The commercial side of scoping is the Associate leading the engagement’s responsibility, not the Senior Associate’s, but it is worth understanding how it works. Most Salamander engagements are priced on a monthly or quarterly retainer that reflects the fractional cadence, with the engagement length and the seniority of the people involved as the principal drivers. Some engagements include a success component tied to a measurable outcome — margin improvement, transaction completion, programme delivery — though these are sized carefully against what Salamander can credibly influence.

The Salamander discipline is to scope at a price that allows the work to be done properly. Engagements that are scoped under-resourced — too little time per week for the work to be delivered to the standard, too few weeks for the outcome to be reached — produce disappointed clients regardless of the technical quality of the work.

Common failure modes.

A few patterns recur.

Scope as topic rather than work. The engagement is scoped as “growth strategy” rather than as a specific outcome to be reached by a specific date. The topic admits expansion; the engagement drifts.

Outcomes that cannot be measured. The scoping document names an outcome but the outcome is not phrased in a way either side can later check. Disagreements at the end of the engagement are predictable.

Deliverables without outcomes. The scoping document lists the artefacts but not what they are in service of. The engagement produces the artefacts and the leadership team is no clearer on what has changed.

Capacity under-scoped. The fractional time commitment is too low for the outcome the engagement has been signed up to deliver. The Associate ends up over-running or the outcome under-delivers.

Dependencies unnamed. The engagement assumes the client will provide data, leadership time, or decisions that are not formally agreed in the scoping document. The engagement stalls when the dependency does not arrive.

For you as a Senior Associate, the practical implication is to learn the scoping discipline through observation early in your tenure. The Associate leading the engagement is responsible for the scoping conversation and document, but you will be working inside the scope they have agreed, and the better you understand how it was built, the better you can spot the cases where scope is drifting and the engagement needs a re-scope. Most engagements that need a re-scope at month three would have benefited from a sharper scoping conversation in week one.

Stakeholder mapping is the engagement-start work of identifying who matters, who decides, and how the company’s particular ownership and governance structure shapes both. Chapter 62 set out the underlying craft — the categories, the technique for getting to the actual decision-maker, the discipline of reading the boardroom. This chapter is about the practical work of doing that mapping at the start of a Salamander engagement, and about the dynamics that differ across the four company types we most commonly work with.

The mapping exercise.

Most Salamander engagements benefit from an explicit stakeholder mapping exercise in the first two weeks. The output is usually a working document — a single page is often enough — that names each stakeholder, their role, their position on the engagement (supportive, neutral, sceptical), the level of engagement the work needs with them, and the cadence on which they should be updated.

The exercise is done with the Associate leading the engagement, not in isolation, and is revisited as the engagement develops. Stakeholders move; new ones surface; some who looked important in week one turn out to be peripheral, and some who were not on the original map become central. A stakeholder map that has not been updated by month two is usually a stakeholder map that has stopped describing the engagement.

A worked example follows. The map below shows what a stakeholder mapping output looks like at the start of an engagement with a PE-backed B2B technology scale-up — the most common Salamander client type and the one that exposes the two-layer dynamic most clearly. Each stakeholder is named, sized by influence, coded by position (supportive, neutral, sceptical), and tagged with the cadence the engagement needs with them. The dashed callout flags the alignment risk the Associate should be testing for: a recommendation that suits the supportive opco view but not the sceptical sponsor view is a recommendation that will be reshaped before it lands. The footer notes how the same map adapts for founder-led, public, and corporate-owned companies — collapse to one column, add an external column, or add a parent layer respectively.

Founder-led companies.

In a founder-led company, the founder is almost always the decision-maker — often more concentrated as the decision-maker than the formal governance structure suggests. Boards exist but defer; investors hold seats but tread carefully; the leadership team has been hired by the founder and often operates with the founder’s voice in their head.

The engagement implications are practical. The founder’s view is the view that matters. The leadership team’s stated positions are usually a function of the founder’s positions; surfacing the team’s independent view, where it exists, requires the kind of one-to-one interview work covered in Chapter 59. The recommendation has to be sold to the founder above all else, even when other stakeholders are formally involved.

A second-order observation: founder-led companies often have informal advisors the founder consults outside the formal structure. Identifying these advisors — and reading their likely position — is part of the stakeholder map even though they will never appear in an organisation chart.

Private equity-backed companies.

A PE-backed company has two layers of stakeholder to map. The operating company — CEO, leadership team, board — is one. The sponsor side — deal partner, operating partner, investment committee — is the other. As Chapter 19 set out, the relationships across these two layers vary, and the engagement reads differently depending on who has commissioned the work and who is paying for it.

The stakeholder mapping discipline in a PE-backed engagement includes naming each side separately, identifying the formal decision-makers and the informal influencers within each, and reading the alignment (or misalignment) between the two. A recommendation that lands cleanly with the operating company but is at odds with the sponsor’s value creation thesis is a recommendation that will be reshaped before it lands; the engagement is better off knowing that early.

Public companies.

A public company adds a third dimension to the stakeholder map: external constituents who shape what the company can and cannot say or do. Public-market investors, analysts, regulators, and the public-disclosure regime constrain how the leadership team can act, particularly on anything material to the share price.

The engagement implications are about confidentiality, materiality, and disclosure rather than about who decides — those questions usually still run through the CEO, CFO, and board. But the Associate working in a public-company engagement has to be calibrated to what can be discussed in what forum, what cannot leave the room, and what the leadership team needs the recommendation to be reshaped around because of an upcoming earnings cycle or disclosure event. Public-company engagements are also the ones where the standard of evidence is highest, because the recommendation may end up reflected in disclosures the company has to defend externally.

Corporate-owned companies.

A B2B technology business owned by a larger corporate parent — either as a subsidiary or as a recently acquired entity in transition — has stakeholder dynamics that combine some of the founder-led and PE-backed patterns with one additional layer: the corporate parent.

The parent’s view is usually carried by one or two people — often a group CFO, a divisional president, or the senior executive sponsoring the acquisition. Their priorities frame the engagement, even where the operating company has its own leadership. The corporate calendar — annual planning, quarterly reviews to the parent, executive reviews — shapes the engagement timeline. The cultural overlay of the parent often shapes what the operating company will or will not agree to.

The mapping discipline is to name the parent’s representatives explicitly, read their position carefully, and design the engagement so the recommendation works for both the operating company and the parent. Recommendations that suit the operating company but not the parent are usually reshaped at the parent’s review; the engagement is better off anticipating that.

Common failure modes.

A few patterns recur.

Mapping the formal structure rather than the actual one. The organisational chart names roles; the actual influence is often distributed differently. Founders who have stepped back into chairman roles but still drive decisions, function heads whose informal influence exceeds their formal authority, board members who carry weight far beyond their title. The map should reflect what is actually true.

Treating all stakeholders as equal. The engagement gives each named stakeholder the same level of engagement and update cadence, when the actual stakes vary substantially. The discipline is to calibrate effort to influence.

Not updating the map. The map from week one is still being used in week ten; the situation has moved. The fix is the discipline of revisiting the map at the engagement’s major checkpoints.

Mapping the explicit stakeholders only. The informal advisors, sponsor operating partners, parent-company representatives, or external voices that materially shape the recommendation have not been named. The fix is to ask, beyond the obvious stakeholders, who else influences the decision.

For you as a Senior Associate, the practical implication is to read the company type early and to do the stakeholder mapping with that type in mind. A founder-led mapping looks different from a PE-backed mapping looks different from a public-company mapping. The mapping discipline is the same; the specific dynamics are not. Reading the company type accurately, and doing the mapping work accordingly, is one of the engagement-start activities that saves the most time over the duration.

The Salamander work product is the set of artefacts a client sees on a Salamander engagement, plus the working artefacts the engagement team produces to do the work. The standard the work product has to meet is the same one set out in Chapter 23 and reinforced throughout this guide — investor-grade, decision-ready, outcome-focused — translated into the specific artefact at hand.

This chapter sets out the artefacts that recur in Salamander engagements, the standard each is held to, and the wider ecosystem of Salamander content an Associate can draw on alongside their own engagement work.

The recurring engagement artefacts.

A few artefacts produce most of the work product on a typical Salamander engagement.

The scoping document, the quotation, the standard terms and conditions, and the cover letter — produced through the online assessment tool covered in Chapter 65, with the Associate’s edit on the scope itself.

The assessment memo or report — usually produced two to four weeks into the engagement, capturing what the assessment phase has found and where the recommendation is pointing. Sometimes a memo, sometimes a deck, depending on the engagement and the audience.

The working models — the financial, GTM, or operating models that support the recommendation. As Chapter 49 set out, these are built lean and driver-based, with the Salamander library of models and templates as a starting point where appropriate.

The interim updates — short written or verbal communications to the leadership team or sponsor through the course of the engagement. Usually weekly or fortnightly, often brief.

The final deliverable — the memo, deck, or implementation plan that anchors the recommendation at the end of the engagement. The form is set by the audience and the use, not by template.

The operating cadence materials — where the engagement includes installing or refreshing the cadence, the artefacts that support it (board pack, monthly business review template, KPI dashboard) are part of the work product.

The wider Salamander content ecosystem.

Beyond the engagement-specific artefacts, Salamander maintains a wider library of content the Associate can draw on. The deck toolkit (Chapter 8) covers recurring scenarios and is the starting point for many engagement decks. The Excel models and templates library (Chapter 49) covers recurring financial structures, mainly financial in nature, and is similarly a starting point rather than a destination.

In addition, Salamander produces two categories of published content that Associates can share with clients where relevant. The white papers are formal, client-facing published pieces covering recurring topics — the kinds of structural questions that come up across many engagements. They are useful as references for clients who want to read the underlying thinking, as conversation-openers in scoping discussions, and as content the client can circulate internally to support a recommendation. The founders’ guides are audience-calibrated to earlier-stage company founders rather than to scaled-up leadership teams; they cover the specific issues founder-led companies face and are useful in scoping conversations with that audience and as practical references the founder can absorb.

Senior Associates should know what is in each of these libraries. The time saved by reaching for an existing white paper instead of writing the same content from scratch is considerable; the same is true for the deck toolkit and the model library. The discipline is to start with what already exists and edit, rather than to build from a blank page out of habit.

The standard the work product has to meet.

A Salamander artefact, whatever its form, has to meet a few standards.

It has to be decision-ready. The reader should be able to act on it without further analysis. This means the answer is at the top, the supporting argument is structured, the evidence is referenced, and the implications for what to do are explicit.

It has to be investor-grade. The numbers reconcile to the client’s source data; the analysis would survive a sceptical reader’s scrutiny; the assumptions are named and calibrated to the confidence the data supports. The standard from Chapter 26 (investor-grade reporting) applies across the work product, not just to the reporting work narrowly defined.

It has to be outcome-focused. The artefact serves the outcome the engagement is signed up to deliver. Artefacts that exist for their own sake, that document effort rather than support a decision, are not work product in the Salamander sense.

It has to read like Salamander. Direct, operator-voiced, calibrated to the audience. The discipline of Chapters 55 through 58 — top-down communication, storylined structure, executive-grade writing, minimum viable visuals — applies to every artefact that leaves the engagement.

The internal review discipline.

Most substantive Salamander deliverables go through an internal review before they go to a client. The review involves the Associate leading the engagement, sometimes a second Associate brought in for an independent read, and the discipline is set out in Chapter 61 (constructive challenge). The review’s job is to ensure the work product meets the standard and to catch the issues that the engagement team has been too close to the work to see.

Senior Associates should expect their work to be reviewed, take the feedback constructively, and over time develop the judgement to self-review with the same eye an experienced Associate would bring. Engagements where the review discipline holds produce work product that travels well; engagements where it slips usually produce work product that surprises the client in ways that should have been caught earlier.

Common failure modes.

A few patterns recur.

Work product that is technically sound but stylistically off. The numbers are right; the storyline is unclear or buried; the headlines are topics rather than answers. The fix is the storylining and communication discipline from Part 7.

Work product that starts from a blank page when an existing artefact would have served. The Associate builds a deck from scratch when a deck toolkit version exists; the engagement loses time and consistency. The fix is to check the library first.

Work product that has been over-engineered. The deck is sixty pages when twelve would have made the point more sharply; the model has hundreds of inputs when fifteen drivers would have done the work. The fix is the minimum-viable-visual and lean-modelling disciplines.

Work product that has not been reviewed. The work has gone to the client without an internal review because the engagement was running tight. The fix is to make time for the review, even briefly, on anything that materially shapes the engagement.

Work product that ignores the wider Salamander library. The Associate has produced something the white papers, founders’ guides, or model library would have supported with a fraction of the effort. The fix is to build the habit of checking what exists before starting to build.

For you as a Senior Associate, the practical implication is to know what is already in the Salamander libraries — the deck toolkit, the model and template library, the white papers, the founders’ guides — and to make the habit of reaching for them when relevant before building new material. The work product the engagement delivers is yours; the building blocks do not all have to be. Salamander runs more efficiently when Associates use the wider content ecosystem rather than producing every artefact from scratch.

Every Salamander engagement runs on a working cadence — the pattern of formal meetings, working sessions, updates, and day-to-day interactions that hold the engagement together over its duration. Setting the cadence well at the start, and maintaining it through the engagement, is one of the practical disciplines that distinguishes engagements that land cleanly from engagements that drift.

This chapter sets out how Salamander Associates design and run engagement cadences, the standard elements that recur, and the calibration that fits the cadence to the engagement’s actual shape.

Setting the cadence at the start.

The cadence is best set during the scoping conversation rather than improvised in week one. The scoping document from Chapter 65 should name the meetings and updates the engagement will run on — the formal steering committee or review meeting, the working session frequency, the day-to-day touchpoints. Setting expectations explicitly avoids the disappointment that comes from a client expecting more contact or more formal review than the engagement is set up to provide.

The cadence should fit the engagement’s shape. A three-month assessment engagement runs on a different cadence than a two-year CFO-as-a-Service. A fractional engagement at one day a week runs on a different cadence than one at three days a week. The discipline is to design the cadence around the engagement’s reality, not against a template.

The formal review meeting.

Most Salamander engagements include a recurring formal review meeting — sometimes called a steering committee, sometimes simply the weekly or fortnightly review. The participants typically include the engagement-defining client leader (CEO, CFO, function head), the Associate leading the engagement, and any other stakeholders the engagement needs to keep aligned.

The meeting has a consistent structure. A brief status update on what has progressed since the last meeting. The substantive work the engagement is currently producing — the assessment finding, the analysis output, the draft recommendation. The decisions or directional choices in front of the client. The next steps and ownership.

The discipline that makes the formal review meeting work is that each meeting moves the engagement forward rather than just reporting on it. A meeting that produces no decision and no next-step commitment is a meeting that has consumed time without earning its place. The “produce decisions, not just discussion” principle from Chapter 60 applies.

Working sessions.

Between formal reviews, most engagements run working sessions — smaller, less structured meetings that produce specific pieces of the work. Sometimes with a function head to walk a data set; sometimes with the leadership team to test a hypothesis; sometimes with a smaller working group to design a specific intervention. The working session is the form most Salamander work actually gets done in.

Working sessions are usually shorter and more frequent than the formal review meeting. They are the working tool of the engagement, not the deliverable forum. The discipline is to keep them focused, time-boxed, and product-oriented — each session named for the output it produces, in the same way Chapter 60 covered for workshop design.

The day-to-day rhythm.

The day-to-day rhythm of a fractional engagement varies with the engagement type and the Associate’s commitment to it. A CFO-as-a-Service Associate operating at three days a week is present in the client’s day-to-day rhythm in a way that an assessment Associate at one day a week is not.

A few practical patterns apply across most engagement shapes. The Associate is reachable by the engagement-defining client leader during agreed hours, on a defined channel (most often a messaging app for short asks, email or a structured update for longer ones). The Associate is in the client’s relevant operating reviews on the days they are present, whether or not those reviews are formally part of the engagement. The Associate’s calendar reflects the engagement transparently — clients should be able to see when the Associate is available and when they are working elsewhere.

For deeper fractional engagements — CFO-as-a-Service, COO-as-a-Service — the Associate is operationally present in the leadership team, attending the operating cadence as if they were a full-time member of it, and contributing to decisions in real time. For lighter engagements, the rhythm is closer to a regular external advisor with a deeper level of access and commitment.

Updates and communication between meetings.

Most engagements run on a written update between formal review meetings. Sometimes a short memo, sometimes a structured email, sometimes a working document the engagement maintains and shares. The discipline is to give the client visibility into what the engagement is doing without producing reporting overhead for the engagement team.

A useful pattern is the weekly update — a short note (often a few paragraphs, sometimes a half-page) covering what progressed, what is in flight, what needs the client’s input or decision, and what is upcoming. Clients consistently find this useful even when the engagement seems to be running smoothly; engagements that go quiet between formal meetings tend to surface anxiety that better communication would prevent.

The end of the engagement.

The cadence at the end of an engagement matters as much as the cadence at the start. A useful Salamander engagement winds down deliberately — a final review meeting that confirms what has been delivered, a written wrap-up that captures the recommendation and the implementation status, and an explicit handover of any artefacts or follow-up actions that sit with the client team after Salamander steps back.

Engagements that simply end — with the last working session followed by silence — leave the client uncertain about whether the engagement is over and whether Salamander is still available for follow-up. The wind-down cadence prevents that.

There is one more wind-down activity that should be part of every engagement that has gone well: asking the client for a short written quote or testimonial Salamander can reference later. The moment to ask is at the close of the engagement, when the value delivered is fresh and the relationship is warm. A short, specific quote from a CEO, CFO, or sponsor who has seen Salamander work is one of the most useful artefacts for the next round of business development — for the website, for scoping conversations with new clients, and for the deal-flow conversations with private capital relationships covered in Chapter 19. If the ask is left until later, the moment usually passes. The discipline is to treat the quote as part of the engagement’s deliverables, not as an afterthought.

Common failure modes.

A few patterns recur.

Cadence that was not set at the start. The first six weeks of the engagement spend time negotiating meetings that should have been agreed in scoping. Time is lost to the negotiation, and the rhythm never quite settles.

Formal reviews that have become status updates. The meeting reports on what has been done but does not produce decisions. The engagement runs on but the work is not being meaningfully challenged.

Working sessions without named outputs. The session has been scheduled because a working session was due, not because a specific piece of work needs to be produced. Output drifts.

Silence between formal meetings. The engagement goes quiet for a week; the client wonders what is happening; anxiety builds. The fix is the written update discipline.

A wind-down that is not deliberate. The engagement ends without a clear close; the client is unsure whether to call Salamander or not on a follow-up question.

For you as a Senior Associate, the practical implication is to read the cadence carefully on every engagement and to be a reliable part of it. Turn up to working sessions prepared. Produce the weekly update on time. Use the formal review meeting to move decisions, not to report status. The cadence is the engagement’s working architecture, and the Associates who run inside it well are the ones the client wants on every engagement that follows.

The final chapter of Part 8 returns to the value that runs through this guide more than any other: we take on outcomes, not just recommendations. Chapter 3 set out the principle; this chapter sets out what it looks like at the engagement level — how the implementation phase actually runs, where the line sits between what Salamander owns and what the client owns, and the disciplines that turn recommendation into delivered change.

Most Salamander engagements have an implementation phase. Some run inside the engagement; some sit alongside; some are explicitly handed over to the client team. The shape varies, but the discipline of driving to outcomes applies across all of them.

What “taking on outcomes” looks like in practice.

At the engagement level, taking on outcomes means a few specific things.

The engagement is scoped to a measurable change in the business, not to a deliverable. As Chapter 65 set out, the scoping document names the outcome the engagement is signed up to produce — margin improved by a defined amount, recurring revenue mix shifted, a new operating cadence installed and running, a managed services model standing on its own feet, an exit-ready set of financials. The deliverables that follow are the path; the outcome is the destination.

The Associate stays engaged through the implementation phase, not just the analysis and recommendation. The fractional model is designed for this — the Associate is operationally present in the client’s cadence during the implementation period, not just at the kickoff and the close.

The success measurement runs against the scoped outcome. At the end of the engagement, the question is whether the change in the business has occurred — not whether the deliverables were produced on time or whether the client was satisfied with the work product.

The handoff from recommendation to implementation.

Most engagements have a moment where the work transitions from “what should we do” to “what are we doing.” The transition is a critical point in the engagement and deserves explicit attention.

A useful Salamander discipline at this transition is the implementation plan. The plan names what is being installed, who is responsible for each piece, what the timeline is, what the early indicators of success look like, and what would tell the team the plan needs to be revised. The plan is sometimes part of the final recommendation document; sometimes it is a separate document produced just before implementation begins. Either way, it is the bridge from analytical work to operational change.

The implementation plan should also name what Salamander is doing through the implementation phase — the cadence the Associate is present at, the working sessions the implementation team will run, the milestones the engagement is tracking against. Without this explicit framing, the client and the engagement team can find themselves with different views of what the engagement is still delivering.

How implementation support works.

Salamander implementation support varies in intensity. At the heaviest end, the engagement runs through implementation with the same Associate continuing in the fractional seat — a CFO-as-a-Service Associate who built the financial discipline recommendation continues to run the finance function while the discipline is installed. At the lighter end, the engagement supports the client’s own team through implementation, with the Associate available for working sessions, escalations, and review without being the executor.

The posture is the one from Chapter 9 (execution-led partnership): inside the problem, hands on the work, alongside the client. The implementation phase is when this posture pays off — when the recommendation hits resistance, the operating reality of the business is more complex than the analysis assumed, and the original plan has to be adapted. The Associate present through implementation can make those adaptations honestly; an engagement that has stepped back at the recommendation cannot.

The line — what we take on and what we do not.

As Chapter 3 set out, Salamander takes on outcomes that are within the operator’s reach. We do not take on outcomes that depend entirely on factors the engagement cannot influence — a market downturn, a competitor’s move, a customer’s procurement cycle, a regulatory change.

The implementation discipline is to be explicit about this line. The implementation plan should name the assumptions the outcome depends on, the exposures that could break it, and the early-warning signals the team is tracking. If the outcome turns on a load-bearing assumption that fails, the engagement adapts — not pretends the original outcome is still in reach. The integrity value (Chapter 7) applies here as much as anywhere else.

There is also a line around what the client owns versus what Salamander owns. Salamander supports the implementation; the client team executes most of it. A CFO-as-a-Service Associate is in the seat; the finance team is doing most of the work. The discipline is to make the ownership explicit — who is doing what, against what timeline, with what authority — so that the implementation does not stall on ambiguity about who is responsible for which piece.

Measuring outcomes versus deliverables.

A Salamander engagement that is driving to outcomes measures itself differently than one that is delivering recommendations. The measurement set focuses on the change in the business: the metric that the outcome was scoped against, the leading indicators that the change is taking hold, the operating cadence that will hold the change in place after Salamander steps back.

The deliverables produced along the way are means, not ends. A team that has produced fifty pages of analysis but has not moved the metric the engagement was scoped against has not delivered the outcome. A team that has moved the metric materially, even if the formal deliverables were minimal, has.

The implementation cadence.

The cadence during the implementation phase is usually more frequent than during the analytical phase. The Associate is in the client’s operating reviews. Implementation working sessions run more often than assessment working sessions. The written updates between formal meetings reflect what has been done that week, what is in flight, and what is needed from the client team to keep the plan moving.

Common failure modes.

A few patterns recur.

Engagements that stop at recommendation. The analysis is done; the recommendation is delivered; the engagement winds down before the implementation phase begins. The change does not happen.

Implementation plans that are wish-lists, not workplans. The plan names the steps but does not assign owners, timelines, or measurement. The implementation drifts because no one is explicitly responsible for any specific piece.

Outcomes that are redefined mid-engagement. The original outcome turns out to be harder than expected; the engagement quietly moves the goalposts to something more achievable. The fix is the integrity discipline — name the change to the outcome explicitly and let the client decide whether to continue against the new scope.

Implementation support that is not in the cadence. The Associate is technically still engaged but is not turning up in the client’s operating reviews or working sessions. The implementation has effectively been handed back without anyone naming it.

Outcomes claimed that were not delivered. The engagement closes with a claim that the outcome has been achieved when the data does not yet support it. The fix is the honest measurement discipline above.

For you as a Senior Associate, the practical implication is to read every engagement against the outcome it is signed up to deliver, not against the deliverables it is producing. Ask, at every working session and every formal review, whether the engagement is moving toward the outcome or just producing artefacts. The Associates whose engagements consistently land are the ones who hold the outcome focus from the first scoping conversation to the final wrap-up — and who treat implementation as part of the work rather than as someone else’s problem after the recommendation has been delivered.

Salamander’s shared drive is the single repository for every Salamander document. It is reached through a login on the Salamander public website and is one of the two internal locations a Senior Associate uses regularly — alongside the intranet dashboard (Chapter 71).

What is on the shared drive.

The shared drive holds the working content the engagement team draws on across every Salamander engagement. The main categories are familiar from earlier chapters of this guide and worth gathering in one place.

The deck toolkit. A substantial number of PowerPoint decks covering recurring scenarios — the kinds of conversations and issues that come up again and again across clients. Senior Associates use these as starting points for engagement decks, openers in scoping conversations, and reference material for the standard ways Salamander has framed common problems.

The Excel models and templates library. An extensive library of models and templates, mainly financial, built up over many engagements. Useful as a starting point for the working models on most Finance engagements — though, as Chapter 49 set out, generative AI has surpassed a meaningful share of what these models were originally built to support. Reach for the library where the structure is sound; rebuild from scratch where AI assistance produces a better answer faster.

White papers. Formal, externally published pieces written for clients and prospects on the structural topics that come up across many engagements. Audience is leadership teams at scaled-up technology companies and the private capital sponsors behind them. Useful as conversation-openers, references for clients who want to read the underlying thinking, and content the client can circulate internally to support a recommendation. Most titles have a matched internal playbook (see below); a few are deliberately white-paper-only — positioning pieces on topics that frame how Salamander thinks rather than engagement-shaped work the firm runs end-to-end. Regional editions exist for several titles (Europe, Middle East, Africa); reach for the regional edition where the client sits.

Founders’ guides. Audience-calibrated to founders of earlier-stage companies rather than to scaled-up leadership teams. External use. The current library focuses on the capital and financing journey — cap table management, term sheets, convertible loan notes, SAFEs, valuation methods, pitch decks, investor updates, and similar — and complements the white papers, which address operating questions that arise once the business is past founder stage. Useful in scoping conversations with founder-led companies and as practical references the founder can absorb between meetings.

Internal-use-only playbooks. The working playbooks that capture how Salamander runs particular kinds of engagements. Strictly internal — never shared with clients, prospects, investors, recruiters, partners, or any party outside Salamander. Each playbook is the operational companion to a same-titled external white paper: where the white paper describes the topic, the playbook contains the Day Zero qualifying signals, the first-call diagnostic, the capability plays, the difficult-conversation scripts, the templates, and the engagement lifecycle that an Associate uses to run the work. The matched-pair structure is deliberate — the white paper carries the framing into a client conversation; the playbook stays inside Salamander. Useful for an Associate stepping into an engagement type they have not run before, and for the internal review discipline covered in Chapter 67.

Assessment frameworks (the diagnostics). The six one-page A4 diagnostics introduced in Chapter 44 and reproduced in Appendix A. Internal use, designed to be used in the first two weeks of an engagement to read the company through a discipline-, model-, or stage-level lens and route into the right playbook. Standalone A4 copies sit alongside the deck toolkit for printing and annotation.

This guide does not list the specific titles in each library. The libraries grow regularly, and a list in print would date quickly. The discipline is to explore the drive at the start of any new engagement and to develop a working knowledge of what is already there before building new material.

How to use the shared drive.

The principle that runs through this guide applies here too: check what already exists before starting from a blank page. Most engagements that go well at Salamander start with someone reaching into the shared drive for a deck, a model, or a paper that has been refined across earlier engagements. The work then customises and edits that material rather than producing it from scratch.

Conventions for adding to the drive.

Senior Associates who produce new material — a deck, a model, a paper — that is likely to be useful on future engagements should add it to the drive, with a clear title, a one-line description, and the engagement context that produced it. The drive is more useful the more deliberately it is maintained; ad hoc additions without context tend to become noise the next Associate has to filter out.

For you as a Senior Associate, the practical implication is simple. Learn what is on the shared drive in your first few weeks at Salamander, reach for it at the start of every engagement, and contribute back when your engagement produces something that will be useful again. The drive is one of the things that makes Salamander efficient. Using it well and adding to it deliberately is part of how every Associate keeps it that way.

The intranet hosts the Salamander Tools Dashboard, the second of the two internal locations a Senior Associate uses regularly. Where the shared drive (Chapter 70) holds documents, the dashboard hosts the tools Salamander has built to do specific pieces of engagement and internal work. Both are reached through a login on the Salamander public website. The dashboard itself is internal-use-only, even though some of the tools inside it produce client-facing outputs.

The dashboard structure.

The dashboard is divided into two sections.

Internal Tools are the tools Senior Associates use to run their own work inside Salamander — the Associate Matcher (matching clients with the right Salamander Associates based on skills, availability, and engagement requirements), the Proposal Cover Letter generator, the Introducer Agreement generator, the Schedule 2 engagement letter, the Expense Claim tool, the cross-engagement Analytics, and the Quick-Start Guide that walks new Associates through the full toolkit.

Client-Facing Tools are the tools that produce client-facing outputs or are used directly in client engagements. They are sub-organised by client type: tools that apply across all client types (such as the NDA Generator and the Assessment Launcher), tools for scale-up B2B tech companies, tools for listed and mature tech companies, and tools for channel partners.

The Assessment Launcher and the triage-to-deep-dive workflow.

The single tool a Senior Associate uses most often at the start of a client engagement is the Assessment Launcher. It is the top-level triage tool, and the workflow it sits inside defines how most engagements are scoped.

The workflow runs in four steps. First, identify the client type — Scale-up B2B tech, Listed or mature tech, or Channel partners. Second, run the triage — a fifteen-minute pass across Finance, GTM, and Operations that ranks the three practice areas by priority and recommends which deep-dive to open next. Third, deploy the recommended deep-dive — the full assessment on the priority practice area, calibrated to the client type. Fourth, the gaps and priorities from the deep-dive inform the engagement scope, which then feeds into the engagement letter, the cover letter, and the rest of the scoping artefacts the dashboard’s internal tools generate.

Three pairs of deep-dive assessments live behind the Launcher — Finance, GTM, and Operations — each available in three client-type variants. The Finance assessment for a scale-up covers financial governance, unit economics, and fundraising readiness; the same assessment for a listed company covers financial performance, governance, capital allocation, and M&A capability; for a channel partner, profitability and margin, cash flow and working capital, and vendor finance and rebates. The GTM and Operations assessments follow the same pattern, with content calibrated by client type.

This structure — triage first, deep-dive second, scope third — is how the substantive analytical work the engagement runs on gets started. It is also why the online assessment tool covered in Chapter 65 matters: the deep-dive output becomes the input to the scoping conversation, and the scoping artefacts (the engagement letter, the cover letter, the terms and conditions) flow from the agreed scope.

Other tools and the wider picture.

Beyond the Assessment Launcher and the deep-dive assessments, the dashboard hosts a growing set of specific tools — calculators, route advisors, readiness assessments, and assessments built around specific recurring engagement questions. The Analytics tool sits across the assessment outputs and produces cross-engagement views: score distributions, category heatmaps, priority-flag patterns, and cross-tool client views the firm uses to refine its own approach over time.

The dashboard will continue to grow as new tools are built. This guide does not try to catalogue every tool; it points to the structure and the workflow inside it. The dashboard’s own Quick-Start Guide is the place to go for the current state.

How to use the dashboard.

The principle is the same as for the shared drive: check what the dashboard can do before building anything from scratch. Most engagements have a faster, cleaner starting point through the dashboard than they would have through a blank page. The Quick-Start Guide on the dashboard is the formal walkthrough and is worth reading early in your tenure.

For you as a Senior Associate, the practical implication is to learn the Assessment Launcher first — it is the tool you will use most. Build a working knowledge of the deep-dive assessments that apply to your most common client types. And know where the rest of the dashboard sits, even if you do not use every tool every week. The dashboard is one of the principal pieces of leverage Salamander brings to an engagement. Used well, it compounds across every engagement an Associate runs.

A note to close.

We are delighted that you are part of Salamander. The work ahead — the engagements you will run, the clients you will work alongside, the colleagues you will build the company with — is more interesting and more consequential than most of us anticipated when we first joined.

One thing this guide has said in several places is worth saying once more at the end. Salamander is not a company that relies on others to bring in the deals. We do it together. The team is the strength of the company, and the network we are building together compounds that strength over time. The engagements we land, the clients we serve well, the outcomes we deliver — none of them are produced by individuals working alone. They are produced by a network of senior practitioners who choose to work together and who back each other up when it matters.

We will grow the business together. We will do things together that would not be possible alone.

The handbook closes here. The work, and the practice, start now.

Discipline-level assessment lens. Source: Chapter 24.

The first read on a Finance engagement: are the five pieces of financial discipline in place, and where is the weakest link? Use as a starting hypothesis, not a checklist.

The five elements. A close that closes on time — predictable monthly close day, material accuracy, reconciliations done, no material movement after close. A rolling forecast that is actually used — thirteen weeks cash, twelve to eighteen months P&L, refreshed monthly, owned by Finance, informed by operating leaders. Controls matched to the size of the business — proportionate to where the risks actually live (rev rec, cash, expenses, payroll, system access), not imported wholesale from a larger company. KPIs that drive behaviour — the right metric set for the model, instrumented to be produced consistently, embedded in the operating cadence. An operating cadence that makes the numbers useful — monthly business review, quarterly forecast update, weekly cash where relevant, a board cycle that pulls it together.

Read each element. Score each as healthy, fragile, or missing. The discipline is a system — close enables forecast, controls enable cadence, KPIs are connective tissue. Building any piece in isolation rarely sticks.

Questions for the first two weeks. What day does the close actually land — and does that day drift? Is the forecast a rolling tool used in operating reviews, or an annual budget reviewed quarterly? Where are controls over-built, under-built, or absent against the actual risks? Which KPIs sit in the operating rhythm, and which only appear as appendices in the board pack? Does the leadership team run the business on the numbers, or only file them? Who in the finance team can run the discipline once installed — and where will a permanent hire be needed?

Common binding constraints. Close drift — unclear ownership, manual tooling, finance team firefighting through the close window. Budget masquerading as forecast — no rolling view; leadership cannot see what is coming. Controls mis-sized — either slowing the business or invisibly fragile at the next diligence. KPIs as decoration — produced but not used. No operating cadence — the numbers exist; the rituals that turn them into decisions do not.

When to set the framework aside. Pre-Series A businesses where the discipline does not yet need to be operating-model grade — meet them where they are. Businesses already at investor-grade discipline, where the Finance work is reporting upgrade or transaction readiness rather than discipline build.

Companion playbooks. The primary companions are the SaaS Metrics and Unit Economics Playbook and the Revenue Operations (RevOps) Playbook. Close drift routes to SaaS Metrics §4.8 (Measurement Infrastructure and Governance). Budget masquerading as forecast routes to SaaS Metrics §4.7 (Forecasting and Scenario Modelling) and RevOps §4.4 (Forecasting and Pipeline Management). Controls mis-sized routes to SaaS Metrics §4.8 and RevOps §4.3 (Data Hygiene and Governance). KPIs as decoration routes to SaaS Metrics §4.1 (Metrics Strategy and Reporting Framework). No operating cadence routes to SaaS Metrics §4.1 and the SaaS Business Model Transformation Playbook §4.8 (Transformation Programme Governance) where the cadence install sits inside a wider operating-model change.

Output. A short written read (memo, not deck) naming the weakest element, the underlying drivers, and the work that follows.

Discipline-level assessment lens. Source: Chapter 29.

The first read on a GTM engagement: walk the funnel end to end, find the discontinuity, and name the binding constraint. The visible symptom is usually not the cause.

The funnel as backbone. Read the company’s GTM as a single funnel: market → awareness → demand → pipeline → conversion → customer → renewal → expansion. The stage names change across motions (human-led, product-led, e-commerce-led); the logic does not.

What to look at in the first two weeks. The data first — pipeline coverage (raw and weighted), stage conversion, cycle time, win rate, CAC and CAC payback, gross retention, NRR, cohorts, expansion economics. Find discontinuities, not benchmarks. The motion next — sit in a pipeline review or a customer success review, listen to sales calls, walk the product onboarding, try to buy something from the website. The motion shows the cause. The team after that — who owns growth at leadership level, what the review cadence is, whether marketing knows what sales is forecasting, whether customer success knows what marketing is promising. The strategy underneath — whether the ICP is actually the customer being won, whether the value proposition lands with that ICP, whether pricing reflects the value delivered.

Recurring binding constraints. ICP drift — winning the wrong customers, with elongated cycles, lower retention, and higher cost-to-serve; symptom shows at conversion or retention. Pipeline starvation — marketing not producing enough qualified pipeline; sales gets blamed but the constraint is upstream. Retention leak — customer success producing more churn than sales can replace; NRR is canonical, drivers are fit, onboarding, reliability, and account management. Motion at the wrong altitude — the motion that worked at the previous size has not been redesigned for the current size. Channel mismatch — leaning on channel to compensate for a broken direct motion, or failing to use channel to amplify a working one.

Questions to keep at the front of your mind. Where does the funnel narrow sharply — and what is upstream of it? Which cohort or segment behaves differently from the headline numbers? If the company has more than one binding constraint, which one should be addressed first?

When to set the framework aside. Pre-revenue product-led testing where the funnel is not yet selling at meaningful scale. Pure transactional e-commerce where the expansion stage is structurally absent. In both cases, retain the principle of finding the binding constraint but adapt the stages.

Companion playbooks. GTM has the deepest playbook coverage in the firm, and the diagnostic is most useful as a router between them. ICP drift routes to the SaaS Go-to-Market Playbook §4.2 (ICP Definition and Market Sizing) and the Enterprise Sales Methodology Playbook §4.1 (Ideal Customer Profile and Territory Design). Pipeline starvation routes to the Marketing for Technology Companies Playbook §4.3 (Demand Generation) and the Pipeline Management Playbook §4.2 (Qualification Discipline). Retention leak routes to the Customer Success Playbook — particularly §4.5 (Renewal Discipline) and §4.6 (Health Scoring and Data). Motion at the wrong altitude routes to SaaS GTM §4.4 (Sales Motion Design) and §4.8 (GTM Organisation and Leadership), and to the Product-Led vs Sales-Led Growth Playbook where the motion question is structural. Channel mismatch routes to the Partner and Channel Strategy Playbook and, where the channel is broken, the When the Channel Goes Wrong Playbook. International growth questions surfaced by the assessment route to the International Expansion Playbook.

Output. A short written memo with one or two supporting exhibits — naming the binding constraint, the underlying drivers, the data supporting it, and the work that follows.

Discipline-level assessment lens. Source: Chapter 37.

The first read on an Operations engagement: assess delivery performance across four dimensions and six functions, with particular attention to the seams between functions.

The four dimensions. Reliability — does the service do what the customer expects, when it expects, at the quality it expects (uptime, SLA adherence, on-time on-scope delivery, response and resolution rates). Productivity — rate of output relative to cost (engineering throughput per developer, tickets per agent, project hours billed vs. planned, MS accounts per ops engineer). Margin — delivery margin by customer, product line, service offering, and PS engagement; most companies lack visibility at the level required to act. Customer experience — what the customer takes away; NPS and CSAT plus the harder signals, including whether the customer feels the company is in command of its own delivery.

The six functions and the seams between them. Product development, product management, professional services, managed services, support services, project management. Each is a delivery motion. Most delivery problems sit either in a single function under-resourced or mis-designed for its work, or — more commonly — in a seam where two functions hand work to each other without discipline.

Assessment order. Customer-facing signals first (reliability data, CX scores, complaint patterns, escalations into account management). Productivity next (output per dollar across the six functions — where improving, stalled, slipped). Margin after that (cost-to-serve across customers, product lines, service offerings). Seams last (handovers between PM and engineering, engineering and support, PS and MS, any of these and the customer).

Common failure modes. Engineering throughput that has not kept pace — usually a question of organisation and prioritisation, not headcount. Professional services that is unprofitable — pricing and project discipline together; fixing either alone rarely works. Support that is overwhelmed — L1/L2/L3 discipline broken; redesign plus shift-left plus AI. Managed services inconsistent across accounts — standardisation is the recurring scope. Seams that drop work — the most common single failure mode.

Don’t miss the structural layer. Automation, AI, and shift-left are reshaping delivery economics. The substantive question is rarely whether to use them, but where — against measurable change in cost-to-serve, cycle time, reliability, or customer experience.

Companion playbooks. Engineering throughput problems route to the Product Development and Management Playbook §4.4 (Delivery Execution) and §4.9 (Organisation, People, and Operations). Professional services that is unprofitable routes to the Professional Services Practice Playbook §4.3 (Pricing and Margin Architecture) and §4.4 (Delivery Methodology and Quality). Support that is overwhelmed routes to the Support Services Playbook §4.3 (Shift-Left Programme and Self-Service) and §4.4 (AI and Automation in Support). Managed services delivery that is inconsistent across accounts routes to the Building a Managed Services Practice Playbook §4.4 (Operating Model and Runbook). Seams that drop work usually route to the Professional Services Playbook §4.7 (Customer Success Integration) and into the relevant function-level playbook. The structural automation, AI, and shift-left layer routes to the AI Strategy for Technology Companies Playbook.

Output. An integrated read across the four dimensions and six functions, with the binding constraint named and the cadence to hold any fix in place specified.

Business-model-level assessment lens. Source: Chapter 20.

The first read on any client: which model is the business actually in, which variant, and is it drifting? The same presenting problem has different causes in different models — and the recommendation that helps one will sometimes hurt the other.

The main models. Pure SaaS — recurring subscription, light implementation, GM typically 70–85%; unit economics turn on CAC payback, gross retention, NRR; dominant disciplines are GTM and product. Services-led software — software with heavy implementation and configuration; mixed GM (software in the 70s, services in the 20s–40s); mix management matters; dominant disciplines are GTM, Operations, and Finance. Managed services — operating a service on behalf of customers; recurring and contractual, closer to operating revenue than subscription; margin depends on operational efficiency; dominant disciplines are Operations, Finance, and GTM (expansion). Hybrid — the most common state; reading the actual mix is part of the early work.

Read three lenses. The revenue ledger — what portion is genuinely recurring, what is project-based, what is one-off licence. The cost base — where the heavy people sit (engineering, services, ops, sales) and how cost-to-serve actually behaves with customer count. The customer base — concentration, contract length, renewal rate, expansion rate. Triangulating across the three usually tells you which model the business is in and where it may be drifting.

Pay particular attention to drift. A business that started life as pure SaaS but has accumulated heavy PS to win enterprise customers is in a different model than its management may still describe it as. A managed services business quietly losing annuity discipline through one-off concessions may be drifting toward project-based. Naming the actual model versus the stated model is often the single most useful early contribution Salamander makes.

What changes when the model changes. The metric set — NRR and CAC payback are the wrong lens for managed services; cost-to-serve, backlog, cohort GM, and renewal rate are right. The growth lever — SaaS turns on the sales engine and product, services-led also on services capacity, managed services on landing customers and expanding scope inside them. The margin profile — SaaS stable and high; services-led moves sharply with mix; MS moves with efficiency, contract structure, and concentration. The transition risk — many of our most valuable engagements sit inside model transitions, most often project-based to managed services or annuity.

Companion playbooks. The model read is what routes the rest of the engagement. A pure-SaaS read routes to the SaaS Go-to-Market Playbook, the SaaS Metrics and Unit Economics Playbook, and the SaaS Pricing Playbook as the core set. A services-led read routes to the Professional Services Practice Playbook alongside SaaS Metrics. A managed services read routes to the Building a Managed Services Practice Playbook and the Support Services Playbook. Where the diagnostic surfaces drift — the business stated to be one model but operating as another — or an active transition between models, the primary companion is the SaaS Business Model Transformation Playbook, particularly §4.1 (Transformation Thesis and Strategic Case), §4.2 (Commercial Model Redesign), and §4.4 (Financial Architecture).

Output. An explicit statement of model, variant, any drift, and the metric set, growth levers, and margin profile that follow. The rest of the work usually flows more cleanly once that foundation is in place.

Business-model-level assessment lens. Source: Chapter 21.

The metrics worth attending to are different across models and stages. The aim is a credible scorecard within the first two weeks — what is healthy, what is misleading on first read, what is flashing warnings.

The five metrics that carry most of the weight. Recurring revenue mix — genuinely recurring vs. project-based vs. one-off; the trajectory matters as much as the level. Gross margin — bedrock of unit economics; watch the trajectory and the variance across cohorts and product lines, because aggregate stability can hide worsening newer cohorts. Retention — gross retention (do customers stay) and NRR (do dollars retained and expanded exceed dollars churned); the most important set in recurring-revenue businesses, and a leading indicator. Profitability — EBIT, operating margin, and Rule of 40 (growth plus operating margin); avoid EBITDA, because D&A is real cost and EBITDA flatters in ways that do not survive scrutiny. Cash — cash on hand, free cash flow, runway, and the burn rate that connects them; the metric most tied to the timing of major decisions.

By discipline. Finance turns on revenue quality and the cash story — recurring mix, rev rec discipline, gross margin and drivers, forecast accuracy, close integrity, runway, FCF. GTM turns on growth efficiency and retention — new ARR at acceptable CAC payback, pipeline coverage and conversion, sales productivity, channel productivity, GR, NRR, expansion. Operations turns on delivery economics and cadence — cost-to-serve, delivery margin, utilisation, delivery-cycle predictability, and the review cadence.

Read in combination, not in isolation. No single metric tells you whether a business is healthy. Growing at 40% with NRR below 95 hides a retention problem inside acquisition strength. High GM with deteriorating mix hides a future margin problem. Strong cash with slowing growth may be optionality, or quiet under-investment. Score against the combination — and against this stage and this model — not against an abstract benchmark.

Slow metrics versus fast metrics. Boards and leadership teams under pressure focus on metrics that move easily — top-line growth, headcount, this quarter’s pipeline — at the expense of the slower metrics that matter more: NRR, cohort gross margin, rolling forecast accuracy. Part of Salamander’s contribution is to keep the slow metrics in the room when the fast ones are dominating the conversation.

Companion playbooks. This diagnostic maps one-to-one onto the SaaS Metrics and Unit Economics Playbook. Recurring revenue mix routes to §4.2 (Revenue Metrics — ARR, MRR, NDR, GRR). Gross margin routes to §4.3 (Unit Economics) and §4.6 (Segmentation and Cohort Analysis). Retention routes to §4.2 and to the Customer Success Playbook in full, since NDR and GRR are CS-driven. Profitability routes to §4.3 and §4.4 (Efficiency Metrics — Magic Number, Sales Efficiency). Cash routes to §4.5 (Cash and Capital Metrics). Where the metric set itself needs to be redesigned to match a different model — the most common Finance-adjacent engagement — the companion is the SaaS Business Model Transformation Playbook §4.4 (Financial Architecture).

Output. A one-page scorecard naming healthy, misleading, and warning metrics — built in two weeks, refined over the engagement.

Stage-level assessment lens. Source: Chapter 22.

Locate the engagement on the stage map before going deep on the work. The failure mode at each stage usually has roots in an unaddressed pressure point from the previous one — reading backwards saves three or four weeks of working on the wrong question.

Around Series A — founders to a business beyond them. The pressure point is the transition from a business that runs on the founders to one that can begin to run beyond them. The recurring failure modes: ICP never sharpened; sales process that does not reliably convert; forecast that will not survive a board; leadership team that is still a founder layer with reports beneath. The typical Salamander move is a fractional CFO-as-a-Service or CRO-as-a-Service while permanent hires are recruited.

Around Series B–C — growing at all to growing predictably. The pressure point is building the infrastructure of a larger company; scaling beyond the founder-led motion. The recurring failure modes are the failures of premature scaling — hiring ahead of the operating model that can absorb the hires; opening a new geography before the home motion is repeatable; building channel before direct is reliable; building a marketing engine that produces volume sales cannot convert. The typical Salamander move is to identify which of these mistakes is the current binding constraint and address that one before the others.

Around the late scale-up — growing fast to growing durably. The pressure point is that investors, customers, and the board are asking for predictability that has not yet been built. The recurring failure modes: unit economics that quietly worsened over six quarters while growth masked them; one or two leadership members the next phase will not survive with; a recurring mix stalled or reversed; a cost base added in good times and hard to unwind. The visible symptom is forecast accuracy; the underlying issues are usually deeper.

Around the established business — defending and reaccelerating. The pressure point is that growth has slowed for reasons no longer obvious — sometimes external, sometimes structural, sometimes the cumulative weight of decisions taken three or four years earlier. The recurring failure modes: attributing slowdown to the wrong cause (sales execution vs. product-market drift; cost vs. revenue mix); transforming without an operating-model change; cutting cost in places that will be needed when the company reaccelerates; tolerating leadership past the point at which the next phase requires different people.

The pattern across stages. Today’s failure mode is usually yesterday’s unaddressed pressure point. ICP not sharpened at Series A becomes a converting problem at Series B. Headcount past the operating model at Series B becomes margin pressure at the late scale-up. No forecasting discipline at the late scale-up becomes a credibility problem at the established stage. Read backwards from the current pressure point to its origin.

Companion playbooks. Stage is a routing lens for the other diagnostics rather than a direct match to a single playbook. Around Series A, the most common companions are the SaaS Go-to-Market Playbook (sharpening ICP and the sales motion), the SaaS Metrics and Unit Economics Playbook (building forecast and board-grade reporting), and the Pipeline Management Playbook (installing qualification and forecast discipline). Around Series B–C, common companions are the SaaS Go-to-Market Playbook, the Revenue Operations Playbook, the International Expansion Playbook, and the Partner and Channel Strategy Playbook — these are the four motions most often scaled prematurely. Around the late scale-up, common companions are the SaaS Metrics Playbook (the predictability rebuild) and the SaaS Business Model Transformation Playbook (when the underlying issues are model rather than execution). Around the established business, common companions are the SaaS Business Model Transformation Playbook, the AI Strategy for Technology Companies Playbook (the structural relevance question), and the When the Channel Goes Wrong Playbook (where channel decisions made years earlier are now visibly broken).

Output. A clear statement of which stage the company is actually in, which pressure points are dominant, which failure modes are visible, and which are still latent.