Fix the Rope First,Then Argue About the Summit
Bill Aulet’s claim is that founding a company is a craft with an order of operations, and that most ventures fail by doing the operations in the wrong order. Twenty-four steps, six questions, two numbers that decide whether a model is a business, and a rule that you keep going back down.
The Twenty-Four Steps,Laid Out in the Order They Are Walked
A companion to the synopsis. Every step of Aulet’s method under the theme it belongs to, with the question it answers, what it produces and the mistake it exists to catch, plus a short set of rules for when to go back down.
1 · OpeningThe rope goes in before the weather turns
Nobody summits a big mountain by inspiration. Weeks before the push, a small team of Sherpas and guides fixes rope along the route, anchor by anchor, and marks the camps. Climbers then go up and down that rope repeatedly, sleeping higher each time, coming back down to recover. The rotations look like wasted effort to anyone watching from base camp. They are the whole method. Bill Aulet’s book makes the same unglamorous claim about starting a company: the route can be fixed in advance, the anchors are known, and the founder who refuses to use them is not brave but under-prepared. Twenty-four anchors and six camps, plus one rule: you keep going back down.
Fifteen minutes end to end, or twenty with all three asides open. Three, if you read Exhibits 1, 3 and 6, then What to carry away.
2 · The argumentEntrepreneurship is a craft, and crafts have sequence
Aulet’s thesis has two halves. The first is that founding an innovation-driven company can be taught; the opening chapter is titled “News Flash: Entrepreneurship Can Be Taught!” and the rest of the book is an attempt to prove it. The second half is the more useful one. If entrepreneurship is a craft, it has an order of operations, and the most common process failure is doing the operations in the wrong order: building the product before knowing who will pay for it, pricing before understanding how the customer decides, raising money before knowing whether the unit economics can ever close.
The 24 steps are that order. They start with a single question, who is the customer, and refuse to leave it until the founder can name a specific market segment, a beachhead inside it, a persona inside that, and the next ten real people who will buy. Only then does the book turn to what the product does, how the customer acquires it, how the company makes money, and, nineteen steps in, how to design and build the thing. The last step is a product plan for the markets after the first one. The sequence is the argument. Everything else is scaffolding for it.
3 · The author’s vantageA practitioner who became the curriculum
Aulet spent eleven years at IBM before running two technology companies, Cambridge Decision Dynamics and SensAble Technologies, and then moved to MIT in 2009 to lead the Martin Trust Center for MIT Entrepreneurship. He holds a professorship of practice at MIT Sloan. The book grew out of the course he teaches there, and the framework is now the centre’s house method, taught to thousands of students and licensed to schools around the world.
A finance reader should treat the framework the way they would treat a consultancy’s methodology: sound and tested, and owned.
4 · Key insightsFour ideas that carry the book
A paying customer is the only condition that counts
Aulet opens step 1 with a claim he calls the single necessary and sufficient condition for a business: someone pays you. Not a technology, not a patent, not a team, not a vision. The claim sounds trivial until you watch what it does to the early steps. Market segmentation stops being a marketing exercise and becomes a search for the group of people most likely to hand over money soonest. Primary market research, which the second edition pushes far harder than the first, exists to find that group in person rather than in a spreadsheet. The CB Insights post-mortems that put “no market need” at the top of the failure list, at 42 per cent, are the mirror image of Aulet’s condition: the companies that died built something before anyone had agreed to pay for it. The 2024 CB Insights update, drawn from a larger set of venture-backed shutdowns, renames the leading root cause as poor product-market fit at 43 per cent and reclassifies running out of capital as the symptom rather than the cause. The renaming makes Aulet’s point more sharply than the original figure did: the money runs out because nobody was buying, and the step that would have found the buyer was never done.
Small enough to dominate, connected enough to leave
The beachhead idea is borrowed from military planning by way of Geoffrey Moore, and Aulet sharpens it into a filter. The first market must be one the company can win outright, because the reference customers, the word of mouth and the sales learning all come from dominance, not presence. It must also lead somewhere. A beachhead with no adjacent segment is a cul-de-sac, however profitable. His working range for a beachhead’s annual revenue potential is deliberately modest, and his step 14 forces the founder to size the markets behind the first one before the first one is even won. Tesla’s 2006 master plan, an $89,000 roadster funding a cheaper saloon funding a volume car, is the shape the book keeps pointing at.
Two numbers decide whether the model is a business
The financial spine of the book is the pairing of lifetime value with the cost of acquiring a customer, and Aulet’s bar is that the first must be at least three times the second, sustained rather than achieved once. Every SaaS board already watches that ratio. What should hold a CFO’s attention is the insistence that both numbers are outputs of earlier steps. LTV falls out of the business model, the pricing framework and retention; COCA falls out of the decision-making unit, the acquisition process and the sales channel. A founder who fills in the two numbers directly has skipped nine steps and will not know which one was wrong when the ratio fails. Aulet is also blunt that founders undercount COCA because they include only the marginal cost of the last sale.
The steps form a loop, not a ladder
The numbering suggests a staircase. The practice is closer to the mountaineer’s rotations. Step 9, identifying the next ten customers, is placed under the customer theme but is done after the product specification, because it is the first hard test of everything above it, and it routinely sends the founder back to step 2. Step 20 asks for a list of the assumptions the whole model rests on; step 21 tests them, cheaply, one at a time; only step 22 permits a minimum viable business product, and step 23 asks whether the customer pays, uses it and tells someone else. Aulet’s phrase for that last test, that the dogs must eat the dog food, is his refusal to let a launch stand in for evidence.
Key. SME: small and medium enterprise, Aulet’s term for a business that serves a local market, pays its founder early, grows in a line and seldom needs outside equity. IDE: innovation-driven enterprise, his term for a business built on an innovation, aimed at a global market, that loses money for years before growing steeply and needs outside capital to get there.
5 · Where the argument echoesThree structures from other fields
The book’s mechanisms rhyme with patterns outside management, and the rhymes sharpen what the book leaves implicit.
In island biogeography, a founding population survives on a small island only if the island is close enough to a mainland or another island for later dispersal. Aulet’s beachhead has the same two conditions, small and connected, but the ecological version explains why the second matters so much: an isolated colony can thrive and still go extinct as a lineage. A profitable niche with no adjacent segment is that colony.
In systems engineering, the V-model puts requirements capture and traceability before any build and ties every test back to a requirement. Steps 6 to 8, the full life cycle use case, the high-level specification and the quantified value proposition, are requirements engineering under another name. The parallel shows why the book can coexist with agile methods rather than oppose them: iteration happens inside each step, but the steps themselves are a V.
In clinical development, phase I trials establish safety on a handful of subjects before phase III establishes efficacy on thousands, and a drug that skips to phase III is a scandal, not a shortcut. Steps 20 to 23 are a phased trial. The minimum viable business product is a dose-finding study, and step 23 is the efficacy endpoint. Seen this way, a launch without a prior assumption test is not bold, it is a protocol violation.
6 · Business parallelsWhere the steps are visible in public companies
None of the companies below followed Aulet’s numbering. Each of them can be read through it, and that reading is the useful part. The table reads each case against the step it illustrates and the step it appears to have skipped.
| Company | The step made visible | What the step looked like in practice | Where the ladder was skipped or reordered | Evidence | Source |
|---|---|---|---|---|---|
| Tesla | Steps 2 and 14 · beachhead and follow-on markets | Enter at the premium end with a roadster, use the cash and the credibility to move down-market in two further models. | Step 19: the cost of acquiring the second and third markets was funded by capital markets rather than by the beachhead alone. | Documented | Tesla blog, 2006 |
| Steps 3 and 5 · end user profile and persona | One campus, one demographic, one use case, before any other university was allowed in. | Step 15: the business model arrived years after the beachhead was won. | Documented | Harvard Crimson, 2004 | |
| Amazon | Step 2 · a beachhead chosen for its leverage | Books as a first category: standard product, deep catalogue, easy to ship, and a customer base that would buy other things. | Step 17: lifetime value was a bet on categories not yet sold, not a number that could be calculated at the time. | Documented | Amazon shareholder letter, 1997 |
| Slack | Step 9 · the next ten customers | A tool built for the founders’ own studio, then handed to a small set of friendly teams before any public launch. | Step 1: segmentation came after the product existed, which is the reverse of the book’s order. | Reported | First Round Review, 2015 |
| Zoom | Step 11 · competitive position on the customer’s terms | Compete on the two attributes the end user actually felt, that a call connects and stays connected, against incumbents competing on feature breadth. | Step 12: the enterprise decision-making unit’s security criterion was learned late, under public scrutiny in 2020. | Documented | Zoom blog, 2020 |
The evidence column is there to stop the table being read as five equal proofs. Four of the five rows rest on a primary document the company itself published — a master plan, a shareholder letter, a founder’s apology — and one rests on a founder’s later account of his own launch, which is the weakest class of evidence in a table like this because it is told by the winner after the outcome is known. Read the Slack row as a plausible reconstruction rather than a record.
One thing stands out in the fourth column. The ladder is most often broken at the money steps, 15 to 19, and broken by companies with enough outside capital to defer the reckoning. That is a comment on the capital markets of the last fifteen years as much as on the founders.
Key. LTV: lifetime value of an acquired customer, the present value of the profit one customer contributes over the whole relationship, after gross margin, churn and the venture’s cost of capital (step 17). COCA: cost of customer acquisition, the total sales and marketing spend in a period divided by the customers won in that period, fully loaded rather than the cost of the last sale alone (step 19).
7 · PrinciplesFour operating rules, expanded and compressed
Expand
Read→ Geoffrey Moore, Crossing the Chasm (1991), for the bowling-pin logic Aulet inherits and tightens.
Try→ Score your current target market against the seven criteria in Exhibit 5 and write down the one you are avoiding.
Ask→ If we won every customer in this segment, what would we be able to do next that we cannot do now?
Connect→ Founder-population dynamics in ecology: colonies survive by dispersal, not by size.
Contract
Principle→ Dominance in a small market beats presence in a large one.
Metaphor→ The first bowling pin is chosen for what stands behind it.
Rule→ Size the beachhead, then size the market behind it, before you sell into either.
Trigger→ When a deck says “total market of $X billion”, ask for the first $20 million.
Expand
Read→ Aswath Damodaran, The Dark Side of Valuation (2nd ed., 2009), on why young-company cash flows must be built from operating drivers.
Try→ Rebuild one venture’s LTV and COCA from steps 12 to 18 and compare with the figures in its investor deck.
Ask→ Which fully loaded sales cost did we leave out because it felt fixed rather than marginal?
Connect→ Driver trees in FP&A: a KPI you cannot decompose is a KPI you cannot fix.
Contract
Principle→ LTV and COCA are outputs of the model, not inputs to it.
Metaphor→ A thermometer reading is not a fever cure.
Rule→ Three to one, sustained across a sales cycle, or the model is not yet a business.
Trigger→ When the ratio is presented without a sales-process map behind it.
Expand
Read→ Eric Ries, The Lean Startup (2011), for the experiment loop Aulet slots into steps 20 to 23.
Try→ List the ten assumptions your plan needs to be true and design a one-week test for the riskiest.
Ask→ Which assumption, if false, makes the other nine irrelevant?
Connect→ Phased clinical trials: safety on a few before efficacy on many.
Contract
Principle→ A build is the most expensive experiment; run it last.
Metaphor→ Check the anchors before you weight the rope.
Rule→ No MVBP until the assumptions it depends on have each been tested once.
Trigger→ When someone proposes a launch to “see what happens”.
Expand
Read→ Peter Drucker, Innovation and Entrepreneurship (1985), for the older distinction between the new business and the innovating one.
Try→ Draw the cash curve you expect for the next five years and label it SME or IDE before you show it to anyone.
Ask→ Are we raising capital because the J-curve demands it, or because it is available?
Connect→ Portfolio theory: a fund and a family business price the same risk differently, and both are right.
Contract
Principle→ The shape of the cash curve is chosen, not discovered.
Metaphor→ A day hike and an expedition pack differently for the same hill.
Rule→ Decide SME or IDE at step 0; everything from team to term sheet follows.
Trigger→ When a plan promises venture returns on a lifestyle-business cost base.
8 · Relevance in 2026When building gets cheap, the customer steps get expensive
Ask what has changed since the first edition in 2013 and the honest answer is the cost of step 22. A working product that once needed a funded engineering team can now be assembled in weeks by two people and a model. That does not weaken Aulet’s argument;The cost curve moved under one step, not all twenty-four.Assembly is the part that collapsed. Prompt-to-app platforms and coding agents have turned a quarter of engineering work into a supervised draft. The figures below are about that quarter alone.Lovable · Stockholm, 2024–25$100m in recurring revenue eight months after launch, with around 45 employees, selling the capability that makes building cheap. Its own beachhead is sharp: people who cannot code but know exactly what they want built.Emergent · Bengaluru, from July 2024Mukund and Madhav Jha’s platform turns plain-language prompts into production software rather than demos, and reached $50m in recurring revenue inside seven months, valued at $1.5bn by mid-2026. Its step 2 is unusually disciplined — small businesses that know precisely what they want and have never been able to afford a developer. Its users are the clearest picture of what cheap assembly permits: millions of applications built with no segmentation at all.Klarna · 2024, then 2025An OpenAI-powered assistant handled 2.3 million chats in month one, on the company’s own account the work of 700 agents. Fifteen months later Klarna was recruiting humans again, its chief executive conceding that a cost-first build had produced lower quality. What broke was not the build. It was the customer understanding beneath it.What this does to the 24 steps.It raises their price relative to everything else. When a product costs a fortnight, the expensive asset is a defensible answer to steps 1 to 5. Cheap assembly also removes the old discipline of scarcity: a team that had to earn an engineering budget was forced to argue for the customer first, and nobody has to argue now. The three-to-one test survives and gets harder, because acquisition costs did not fall with build costs. The platforms selling cheap building are disciplined about steps 1 to 5; most of what gets built on them is not.Sources: TechCrunch, 2025; TechCrunch, 2025; YourStory, 2026; Klarna, 2024; Customer Experience Dive, 2025. it inverts the founder’s instinct about where the effort goes. When the build is cheap, the temptation is to build first and segment later, and a good share of the thin wrapper products of the last few years, built in a fortnight on somebody else’s model, were step-22 outputs with no step-2 behind them.
The book also travels well into the corporate innovation function, which is where many of this site’s readers sit. An internal venture inside a bank or a conglomerate has the same six questions and is usually worse at the first one, because the “customer” is a business unit head rather than a person who pays. Aulet’s insistence on a persona with a name and a diary is a corrective for that, and the DE canvas in the second edition is a usable one-page artefact for a stage-gate committee. The place the framework needs adapting is capital: an intrapreneur does not face the J-curve trough as a survival problem, and so tends to under-test.
Two other books mark the edges of what the 24 steps carry. Paul Cheek’s Startup Tactics (Wiley, 2024), the practice layer beneath this one, supplies the tooling each step needs to become a week of actual work: the outreach sequence, the landing-page test, the instrumented experiment. Aulet tells the founder which question to answer next; Cheek tells them what to do on Tuesday morning to answer it. A reader who finds the 24 steps convincing and still cannot start usually needs the second book, not a closer reading of the first.
9 · Where the argument strainsFour lenses on a framework that resists them
Aulet’s method invites engagement from the disciplines it borrows from. Four lenses bring its genuine claims into sharper relief.
Through the lens of effectuation research, Saras Sarasvathy’s studies of expert entrepreneurs found that they reason from the means they already hold towards whatever goals those means allow, rather than from a chosen goal backward to the means. The 24 steps are causal reasoning in its purest form: fix the customer, derive everything else. Sarasvathy’s subjects would recognise steps 20 to 23 and be puzzled by steps 1 to 5. The tension is real and the book does not address it, though step 0 in the second edition, on starting from a passion, an idea or a technology, is a partial concession.
Through the lens of decision science, the planning fallacy predicts that a founder asked to calculate a beachhead TAM, a lifetime value and a cost of acquisition will produce confident figures that are wrong in the optimistic direction. Aulet knows this, which is why the ratio has to be three to one rather than one to one. But a framework that asks for twenty-four numbered outputs risks manufacturing precision. The reader’s best defence is the one the book supplies half-heartedly: treat every figure before step 23 as a hypothesis with a confidence interval, not a result.
Through the lens of complex adaptive systems, markets co-evolve with the entrants that arrive in them. A beachhead chosen at step 2 changes shape once the company is in it; competitors respond, the persona learns, the adjacent segment moves. The framework treats the sequence as if the terrain held still while the founder mapped it. Aulet’s instruction to keep revising earlier steps is the right answer, and it is under-specified: the book never says how often, or on what signal, a step-2 decision should be reopened. Rita McGrath’s inflection-point machinery, set out in Seeing Around Corners (Houghton Mifflin Harcourt, 2019),The 24 steps have no clock.Aulet says to go back and revise; he never says what should send you back. McGrath’s answer is that inflections arrive gradually and then suddenly, and that the early evidence shows up at the edges of an organisation rather than in its reporting line. Three edges matter for a step-2 decision.1 · The wrong customers start succeedingUsers outside the chosen segment adopt the product and stay, while the beachhead persona needs persuading. Slack began as the internal tool of a failing games company; the signal that the tool was the business came from the people using it rather than from the market it was built for.2 · The decision-making unit changes shapeThe segment holds still while the buying committee does not: a champion is promoted, procurement acquires a veto, a new compliance seat appears. Step 12 was answered for a committee that no longer exists, and steps 18 and 19 quietly stop being true.3 · The ratio drifts with no campaign to explain itAcquisition cost creeps up or retention slips while nothing internal has changed. A ratio that moves for reasons the company did not cause is a statement about the terrain, not about execution.The rule this implies.Put a standing date on step 2 rather than waiting for a crisis to force it — quarterly for a young company — and write the trigger down in advance: which of the three signals, at what magnitude, reopens the beachhead question. A week spent reopening a decision that turns out to be sound is cheap. A year spent executing twenty-two steps against a segment that moved is not.Sources: McGrath, Seeing Around Corners (Houghton Mifflin Harcourt, 2019), on inflection points and peripheral signals; First Round Review, 2015. supplies exactly that missing trigger: the early warnings that tell a founder the beachhead they chose has stopped being the beachhead they are in.
Through the lens of platform economics, the two decisive numbers assume a customer who is both user and payer. On a platform, the two decisive numbers stop being two.Aulet’s money steps run on one customer who uses the product and pays for it. A marketplace has at least two, and the economics of the pair are not the sum of the economics of each. Three assumptions break, in order.1 · Which side pays is designed, not discoveredRochet and Tirole’s work on two-sided markets makes the price structure — how the total charge is split between sides — as decisive as the price level. One side is usually subsidised so the other becomes worth joining. Step 15 asks how you make money; a platform founder must first answer which side is bought and which side pays, and those are different questions with different evidence behind them.2 · Acquisition on one side is product spend on the otherThe cost of recruiting supply is not a supply-side cost. It is what makes the demand side’s product exist at all, which is why early marketplaces do unscalable things — photographing listings, hand-recruiting restaurants, seeding drivers — and book it as marketing. Computed per side, the three-to-one ratio flatters one column and condemns the other. The only honest unit is the matched transaction: what a completed match earns over its life against what both sides cost to bring to it.3 · The take rate is contestableA platform’s price can be reset by someone who is not its customer. Apple’s 30 per cent App Store commission survived the 2021 Epic ruling on antitrust grounds but lost its anti-steering protection, and the fight over what the company may charge on purchases made outside the app has run through the courts ever since. Step 16 treats pricing as a framework the founder sets. On a platform it is a framework the founder proposes and a regulator, a court or the other side’s alternatives may revise.What this does to step 2.It doubles it. A platform beachhead is two segments chosen together and bounded by the same geography or context — these hosts and these guests, this city, this season — because a segment that is winnable on one side and unreachable on the other is not a beachhead. Aulet’s discipline holds; it simply has to be run twice, and the two answers have to be true at the same time.Sources: Rochet and Tirole, Platform Competition in Two-Sided Markets, Journal of the European Economic Association, 2003; Fenwick, 2021. In a two-sided market, lifetime value on one side is a function of acquisition cost on the other, and the three-to-one rule dissolves into a coupled system. The second edition acknowledges primary and secondary customers and two-sided markets inside step 1, but the money steps still run on a single-customer model. A reader building a marketplace has to do that extension alone.
Refracted through these lenses, the critiques above are not friction against Aulet’s argument — they are scaffolding the reader can build around it: places to anchor their own engagement and test the framework against the disciplines it touches. What Aulet built survives this engagement; what the reader gains is a firmer grip on why it does. What survives all four lenses is the sequence claim itself: no lens disputes that the customer steps must precede the money steps, or that the money steps must precede the build. What the lenses expose is where the sequence needs extending rather than replacing — a means-first entry for founders who start from what they already hold, a signal for when a step-2 decision should be reopened, and a second ledger for the side of a platform that does not pay.
10 · Adjacent readingWhere to go from here
Neighbours · same territory, different angle
- Steve Blank and Bob Dorf, The Startup Owner’s Manual (2012), the customer-development process Aulet’s early steps are in conversation with.
- Eric Ries, The Lean Startup (2011), the build-measure-learn loop that lives inside Aulet’s steps 20 to 23.
- Paul Cheek, Disciplined Entrepreneurship Startup Tactics (2024), the companion volume on executing the steps with a small team.
Productive adversaries · books that complicate the argument
- Saras Sarasvathy, Effectuation (2008), the empirical case that expert founders do not start from a chosen market.
- Peter Thiel with Blake Masters, Zero to One (2014), a sceptic of incremental testing who nonetheless shares Aulet’s taste for small markets you can monopolise.
Deeper roots · where the ideas come from
- Geoffrey Moore, Crossing the Chasm (1991), the origin of the beachhead and the bowling-pin sequence.
- Peter Drucker, Innovation and Entrepreneurship (1985), the first serious claim that innovation is a discipline with a method.
11 · ClosingBack down to the camp below
The climber who has been up to Camp III and back twice does not feel like a hero. She feels tired, and slightly foolish to be walking downhill again with the summit in view. Aulet’s book asks the founder to feel that way on purpose: to go back to step 2 after step 9 has embarrassed it, to rebuild the sales process after the acquisition cost has come in double, to treat the minimum viable product as a rotation rather than a push. The fixed rope is not a guarantee. Plenty of well-roped climbers turn back. What it does is make the decision to turn back, or to go on, an informed one, taken at an anchor rather than in the weather.
Which of your current ventures, corporate or personal, has a product but no step 2?
12 · What to carry awayWhat survives a week after reading
Sequence is the thesis. The 24 steps are not a checklist of good practices; they are a claim about which decisions must precede which others.
The customer condition is a filter, not a slogan. Requiring a paying customer at step 1 turns segmentation into a hunt for the first cheque.
The beachhead has two properties, and the second is the one founders forget. Small enough to dominate, connected enough to lead somewhere.
LTV and COCA are outputs. Both fall out of nine earlier steps, which is why a founder who asserts them has skipped the work without noticing.
The framework covers the customer-and-money half of a venture. Effectuation, co-evolving markets and platform economics mark the edges where the reader must extend it.
13 · AbbreviationsTerms used in this synopsis
Short forms used above, expanded, with what each carries in the argument.
| Short form | Expansion | What it means here |
|---|---|---|
| LTV | Lifetime Value | What one acquired customer is worth across the whole relationship. In Aulet’s scheme it is an output of the business model, pricing and retention steps, never an opening assumption (Exhibit 4). |
| COCA | Cost of Customer Acquisition | What it costs to win that customer, fully loaded rather than marginal. Falls out of the decision-making unit, the acquisition process and the sales channel; the book’s test is LTV at least three times COCA (Exhibit 4). |
| TAM | Total Addressable Market | Sized here for the beachhead alone at step 4, and deliberately small: the number that shows a market can be dominated, not one that impresses an investor (Exhibit 1). |
| MVBP | Minimum Viable Business Product | The smallest thing a customer will pay for that also proves the business, not just the product. Step 22, and the point at which building is finally permitted (Exhibit 6). |
| SME | Small and Medium Enterprise | Aulet’s name for the founding path that pays its owner from year one, grows linearly and needs no outside capital (Exhibit 2). |
| IDE | Innovation-Driven Enterprise | The other path: global ambition, a J-curve trough funded by outside money, returns only if the beachhead converts. The 24 steps are written for this one (Exhibit 2). |
| FP&A | Financial Planning and Analysis | The corporate function where the book’s logic transfers most directly, through driver trees that decompose a target into the steps that produce it. |
| KPI | Key Performance Indicator | Used here as the thing a driver tree takes apart: a measure you cannot decompose into the decisions behind it is one you cannot fix. |
| DE | Disciplined Entrepreneurship | The method itself, and the one-page DE canvas added in the second edition that holds all 24 steps on a single sheet for a stage-gate review. |
| DMU | Decision-Making Unit | Everyone besides the user who must say yes — champion, economic buyer, influencers, and whoever holds a veto. Mapping it at step 12 is what makes COCA computable (Addendum). |
| MVP | Minimum Viable Product | The lean-startup term Aulet deliberately amends. An MVP proves that something works; his MVBP proves that someone pays (Addendum). |
1 · About this addendumTwenty-four anchors, in walking order
This page is the companion to the synopsis of Bill Aulet’s Disciplined Entrepreneurship. The synopsis argues about the framework; this addendum simply lays it out. Each of the 24 steps appears under the theme Aulet assigns it to, with four things stated as plainly as possible: the question the step answers, the artefact it produces, and the failure it was written to catch. The tables are a reading of the second edition (Wiley, 2024), not a reproduction of it, and the failures in the last column are the ones most visible from a finance seat.
Two notes on reading. First, the numbering is the order of work, but the themes are the logic, so steps 9, 14 and 18 sit away from their numerical neighbours. Second, Aulet is explicit that later steps overturn earlier ones, and that going back is the method working rather than failing. Where a step is a frequent cause of backtracking, the table says so.
2 · Before the stepsStep 0, the two games, and the canvas
The second edition adds material before step 1 that the first edition scattered across its introduction. Three pieces matter.
Step 0 · Getting started
Aulet names three ways a venture begins: from a passion, from an idea, or from a technology. Only the second and third are ready for step 1. A founder with a passion but no idea is told to go and find one, usually by working inside the domain, before segmenting anything. He also puts the founding team here rather than later, on the grounds that the steps are a team sport and a solo founder will be unable to run the interviews the early steps demand.
SME or IDE
A small-and-medium enterprise serves a local market, pays its founder early, grows linearly and rarely needs outside equity. An innovation-driven enterprise addresses a global market, loses money for years, grows on a J-curve, and needs capital to cross the trough. The 24 steps are written for the second kind. A founder who applies them to the first will over-engineer; a founder who applies SME instincts to an IDE will under-fund.
The DE canvas
The new edition supplies a one-page canvas for tracking which steps have been completed, which are in draft and which were reopened. For a corporate venture committee it is the most immediately usable artefact in the book: a single sheet that shows whether the money steps have been done, or merely asserted.
3 · Theme oneWho is your customer?
Six steps, and the ones most often skipped. The theme runs from a wide segmentation to a single named person and then back out to ten more like them. Its output is not a market description but a decision about whom to serve first and the evidence that those people exist.
| Step | The question it answers | What you produce | The failure it is built to catch |
|---|---|---|---|
| 1 · Market segmentation | Who could plausibly pay for this, and what do the candidate groups look like? | A long list of end-user segments, each described by the same handful of attributes, narrowed to six to twelve worth researching in person. | Falling in love with a technology and back-fitting a customer. Also the opposite: a “customer” defined so broadly (“SMEs”, “hospitals”) that no one in particular buys. |
| 2 · Select a beachhead market | Which single segment can we dominate first, and does winning it lead somewhere? | One chosen segment, scored against the seven criteria (well-funded, accessible, compelling reason to buy, whole product, competition, leverage, values), plus the reasons the runners-up were parked. | Choosing the largest segment rather than the winnable one. Choosing a segment with no adjacent market to move into. |
| 3 · Build an end user profile | Who, in that segment, actually uses the product day to day? | A demographic and behavioural sketch of the end user, distinct from the economic buyer, with the founders honest about whether anyone on the team fits it. | Building for the buyer and forgetting the user, or building for yourself and assuming the user is you. |
| 4 · Calculate the beachhead TAM | If we won every end user in the beachhead, how much revenue a year is that? | A bottom-up number: end users counted, multiplied by an annual revenue per user, checked top-down. Aulet’s comfortable range is tens of millions a year, not billions. | A TAM copied from an analyst report. A beachhead too big to dominate or too small to justify the trough. |
| 5 · Profile the persona | What is the one real person we are building for called, and what is their week like? | A single named individual from the beachhead, with their priorities ranked, their purchasing habits, what they read and who they trust, based on interviews rather than imagination. | The composite persona that offends no one and guides nothing. A persona whose top priority is not the problem you solve. |
| 9 · Identify your next 10 customers | Are there ten more people like the persona, and do they agree with what we have written down? | A list of ten named prospects beyond the persona, each contacted, with their reactions to the use case and the specification recorded, and the persona revised where they disagree. | One enthusiastic early contact standing in for a market. This is the step that most often sends the founder back to step 2. |
4 · Theme twoWhat can you do for your customer?
Five steps that convert the persona into a product description, a quantified benefit, a defensible core and an honest competitive map. Nothing is built yet. Aulet’s discipline here is that the product is described entirely in the customer’s vocabulary.
| Step | The question it answers | What you produce | The failure it is built to catch |
|---|---|---|---|
| 6 · Full life cycle use case | How does the persona find, evaluate, buy, install, use, pay for, and recommend the product, start to finish? | A step-by-step narrative or diagram of the whole life cycle, from the moment the need is felt to the moment the customer tells a colleague. | Designing the moment of use and ignoring procurement, onboarding and support, which is where enterprise sales stall. |
| 7 · High-level product specification | What does the product look like from the customer’s side, without engineering detail? | A visual, brochure-level description of features and benefits, written for the persona rather than the engineers. Aulet suggests literally drafting the brochure. | A feature list that grows with every conversation. A specification the customer cannot picture. |
| 8 · Quantify the value proposition | In the persona’s own top priority, how much better is life with the product than without it? | An as-is versus possible-state comparison expressed in the unit the persona cares about (hours, money, risk, throughput), tied to the top item on their priority list from step 5. | Claiming value in a currency the customer does not use. Quantifying a benefit that is real but third on their list. |
| 10 · Define your core | What will we be able to do, in time, that no one else can, and that customers will pay for? | A short statement of the one capability the company will protect and build around: a network effect, a unique process, a relationship, a proprietary asset. Aulet is explicit that first-mover advantage and locked-up suppliers are usually not a core. | Confusing core with competitive position, or with intellectual property that anyone can design around. |
| 11 · Chart your competitive position | On the two attributes the persona cares about most, where do we sit against the alternatives, including doing nothing? | A two-by-two with the persona’s top two priorities as axes, the company in the top-right corner, competitors placed where they actually sit, and the status quo plotted as the toughest rival. | Choosing axes on which you win rather than axes the customer uses. Leaving the status quo off the chart. |
5 · Theme threeHow does your customer acquire your product?
Three steps that most technical founders treat as somebody else’s problem. Aulet places them before the money steps because acquisition cost cannot be estimated without them, and a plan that guesses at the sales cycle guesses at everything downstream.
| Step | The question it answers | What you produce | The failure it is built to catch |
|---|---|---|---|
| 12 · Determine the decision-making unit | Who, besides the user, has to say yes, and who can say no? | A map of the DMU: champion, primary economic buyer, end user, influencers, and veto holders, with names where possible for the next ten customers. | Selling to the user and losing to procurement, security, legal or a spouse. |
| 13 · Map the process to acquire a paying customer | What does the customer have to do, in what order, over how long, to buy? | A timeline of the customer’s purchase process, budget cycles, approvals and compliance steps, with an estimate of sales-cycle length that later feeds COCA. | A sales forecast that assumes a decision the customer cannot make in the quarter. |
| 18 · Map the sales process | How will we reach and close customers now, in a year, and at scale? | Three sales-channel designs for three horizons: an early direct model, a transition model, and a long-term model with channels and lower cost, each with the DMU and process from steps 12 and 13 behind it. | Planning one channel for all time. Assuming a direct sales force can be afforded at scale, or a channel can be used before the product is understood. |
6 · Theme fourHow do you make money off your product?
Four steps, two numbers. The business model and the pricing framework produce lifetime value; the sales process from theme three produces the cost of acquisition. The test is a sustained ratio of at least three to one. Exhibit B shows where each number comes from, which is the reason the steps resist being skipped.
| Step | The question it answers | What you produce | The failure it is built to catch |
|---|---|---|---|
| 15 · Design a business model | How do we capture a share of the value we quantified at step 8? | A chosen model from a wide menu (one-time, subscription, usage, licensing, consumables, transaction fees, advertising and so on), justified against the persona’s buying habits and the DMU, and reviewed as a strategic choice rather than a default. | Copying the incumbent’s model. Choosing subscription because everyone does, without asking whether the DMU can approve recurring spend. |
| 16 · Set your pricing framework | What is the price band, and on what basis will we move within it? | A framework rather than a number: the value captured relative to step 8, the alternatives from step 11, the DMU’s reference points, and a rule for early-customer discounts that does not anchor the market. | Cost-plus pricing on a product whose value is many multiples of cost. Discounting the first customers so deeply that the price can never recover. |
| 17 · Calculate the lifetime value | What is the present value of the profit one acquired customer contributes over the life of the relationship? | An LTV built from the business model, price, gross margin, retention, upsell and a discount rate that reflects the venture’s risk, with the assumptions listed beside the number. | Using revenue instead of profit. Ignoring churn. Discounting at a corporate rate when the venture’s cost of capital is far higher. |
| 19 · Calculate the cost of customer acquisition | What does it truly cost, fully loaded, to win one customer through the sales process in step 18? | Total sales and marketing spend in a period divided by customers won in that period, calculated separately for each of the three sales-process horizons, and compared with LTV. Aulet wants LTV at least three times COCA. | Counting only the marginal cost of the last sale. Assuming COCA falls automatically with volume rather than because the channel was redesigned. |
7 · Theme fiveHow do you design and build your product?
This is where the lean startup’s experiment loop lives inside Aulet’s sequence. The difference is that the assumptions being tested were generated by nineteen earlier steps, so the founder knows what a failed test would send them back to revise.
| Step | The question it answers | What you produce | The failure it is built to catch |
|---|---|---|---|
| 20 · Identify key assumptions | What has to be true for the previous nineteen steps to hold? | A prioritised list of the assumptions the model rests on, drawn from every earlier step: that the persona will pay this price, that the DMU can approve it, that the channel can reach them, that the core can be built. | Treating earlier outputs as facts. Testing the product before testing the assumptions the product depends on. |
| 21 · Test key assumptions | What is the cheapest experiment that could show each assumption to be false? | One designed test per high-priority assumption, run with real customers, with results recorded and earlier steps revised where the test failed. | Running one big test that confounds ten assumptions. Asking customers what they would do instead of watching what they do. |
| 22 · Define the minimum viable business product | What is the smallest thing a customer will pay for that also proves the business, not just the product? | An MVBP specification: the fewest features that deliver the value from step 8, are sold under the model from step 15, and generate real feedback and revenue. Aulet’s addition of “business” to the familiar MVP is deliberate. | A demo dressed as a product. A free pilot that proves usage but not willingness to pay. |
| 23 · Show that the dogs will eat the dog food | Do customers pay, use it, and tell others? | Evidence from the MVBP on three counts: the customer paid, the customer uses it (with usage data), and the customer recommended it. Plus the revisions to earlier steps that the evidence forced. | Launching and declaring victory. Counting sign-ups as customers. |
8 · Theme sixHow do you scale your business?
Two steps, one of them numbered fourteen. Aulet wants the follow-on markets sized early, while the beachhead is still being fought for, so that the product plan at step 24 is a sequence of markets rather than a list of features.
| Step | The question it answers | What you produce | The failure it is built to catch |
|---|---|---|---|
| 14 · Calculate the TAM for follow-on markets | Once the beachhead is won, where next, and how big is that? | A ranked set of follow-on markets, each sized bottom-up, reached either by selling the same product to an adjacent segment or by selling a new product to the same persona. Aulet suggests the follow-on markets can be an order of magnitude larger than the beachhead. | A beachhead that is a dead end. A follow-on market that needs a different core. |
| 24 · Develop a product plan | What features and products come after the MVBP, in what order, for which market? | A roadmap that sequences features for the beachhead first, then the products and adaptations needed for the follow-on markets, with the assumptions each stage depends on. | A roadmap driven by the loudest customer or the most interesting engineering problem rather than by the market sequence. |
9 · Using the sequenceWhen to go back, and how far
The book is clearer about the order of the steps than about the rules for reopening them. Reading across the 24, four patterns of backtracking recur, and a venture team can adopt them as explicit triggers.
| Trigger | What it usually means | Where to return |
|---|---|---|
| Fewer than half of the next ten customers agree with the use case (step 9) | The persona is real but the problem ranked lower than assumed, or the segment is wrong. | Step 5 first; if the persona survives, step 2. |
| The DMU has a veto holder the persona never mentioned (step 12) | The full life cycle use case stopped at the user. | Step 6, then step 13. |
| COCA comes in above one-third of LTV in the early sales model (step 19) | Either the channel is wrong for the DMU or the price is capturing too little of the value. | Step 18 first; if the channel is right, step 16 and step 8. |
| Customers use the MVBP but will not pay for it (step 23) | The value is real but sits in a currency the DMU cannot spend, or the business model asks for approval the DMU cannot give. | Step 15, then step 8. |
A useful house rule for a stage-gate committee follows from the table: no funding decision on a step-22 build until the DE canvas shows steps 12, 13, 17 and 19 completed with real names and real figures behind them. The step numbers make the rule easy to state and hard to argue with.
The question, artefact and failure columns are the author’s synthesis of each step’s stated purpose and are written for a finance and strategy reader; they paraphrase the book and do not quote it.