Lean Canvas
beta · Family strategy-docs · Phase undefined · Sizes lean · ~6,250 tokens
The one-page record of a product’s business model as beliefs to test, not a finished plan: who has the problem, what value is offered, how it reaches and earns from customers, what it costs, and the riskiest belief to check next. Not a business plan, not a financial model, and not the Business Model Canvas it adapts.
An operator card. Read it before you sketch the canvas, and again before you share it with the first
other person. The reasoning behind every rule here lives in
lean-canvas_companion.md.
When to use this document
Section titled “When to use this document”Sketch a Lean Canvas when the business model behind a product is still a set of beliefs rather than a settled fact: who has the problem, who will pay, how the product reaches them, how it makes money, and what would have to be true for it to win. Ash Maurya built the canvas to “deconstruct an idea into key assumptions or beliefs”, and every box exists to state one of those beliefs in a form that can be tested, not to describe a plan that has already been decided. Write the first sketch alone, even if you work on a team: “If you’re part of a team, avoid creating a Lean Canvas as a group exercise. Instead, have each team member create their snapshot first.” Keep it fast: “your initial big idea canvas should be sketched quickly - in less than 20 minutes”, on one page, with “Set a timer for 20 minutes” as a literal instruction rather than a figure of speech. Share the finished sketch at once: “The important thing is sharing it with at least one other person when you are done.”
Maurya’s own company draws the line for when this document applies by stage, not by company size or
age: “If you are pre-product-market fit, start with Lean Canvas”. This library places the canvas second
in its strategy-docs chain, between the product vision and the product strategy, and names that
placement its own position rather than a sourced finding: no source compares the two sequences by name.
What the placement does not change is what the canvas is for: it exists so a later product strategy has
beliefs to commit to or reject, not the other way around. See lean-canvas_companion.md section 6
(Debates) and section 8 (Relationships) before treating the order as settled either way.
Is this the right document at all?
Section titled “Is this the right document at all?”| If the real question is | Write this instead |
|---|---|
| Where are we trying to get to, over years, and why is it worth it? | product-vision |
| Which problems will we solve to get there, and which will we not? | product-strategy |
| How do we stand against one named alternative, for one customer, now? | positioning-statement |
| Which single measure of customer value shows the strategy is working? | north-star-metrics |
| Does this one investment decision justify the spend? | business-case |
| What does this one customer type actually look like, researched in depth? | user-persona |
| What are the hypotheses behind one initiative, not the whole business? | lean-ux-canvas |
| Already past product-market fit, and want the tool built for an established business model? | the Business Model Canvas, catalog entry 4, not yet built in this library |
When NOT to use this document
Section titled “When NOT to use this document”- The business already has product-market fit. Maurya’s company states the test directly: “the better tool for mapping and communicating established business models” is the Business Model Canvas, not this one, and “If you are pre-product-market fit, start with Lean Canvas” names the opposite case as this document’s own ground. That same source also softens the line, “Use what is most natural for you”, so treat the stage test as the sourced default, not an absolute rule.
- A team is about to use it as a group whiteboard exercise for a first draft. The current teaching page names this directly as the wrong way to start: “If you’re part of a team, avoid creating a Lean Canvas as a group exercise. Instead, have each team member create their snapshot first.” Reconciling individual snapshots afterward is the method; drafting by committee from a blank page is not.
- The problems and customer segment have not been named yet at all. This document asks you to state beliefs about a segment and its problems, not to discover the segment from scratch; if nothing is known yet, talk to customers first and come back to the canvas once there is a first guess to write down.
- The content is a financial model or a multi-year budget. “Rather than thinking in terms of three-
or five-year forecasts, it’s better to take a more stage-based approach.” A real financial model belongs
in
business-case, not in this canvas’s Revenue Streams or Cost Structure boxes. - The content is one initiative’s hypotheses, not the whole business. A Lean UX Canvas frames one piece of work; this canvas states the business model that work sits inside, including revenue, cost and channels, which a single initiative does not need to answer on its own.
- The reader needs a finished plan rather than a current best guess. “Unlike a business plan, the goal with a Lean Canvas isn’t to achieve perfection but to take a snapshot.” A reader expecting a business plan’s completeness will be disappointed by design, not by a defect in the fill.
Picking a variant
Section titled “Picking a variant”There is nothing to pick. This bundle ships one size, lean, and every source that addresses length
agrees the canvas is meant to be compact rather than offered in a longer form. “Embrace a 1-page
constraint”, and the reason given is not a formatting preference: “Describing your idea on a single
page is not about using smaller fonts but fewer words.” The 2011 walkthrough agrees from its own angle:
“Sketch a canvas in one sitting”, and “The boxes on the Lean Canvas are intentionally small because it
forces you to be concise.” It is normal, not a sign of an unfinished draft, to leave a box short: “it’s
okay to leave boxes blank”, and “It’s okay to leave sections blank”. Do not build a longer variant by
copying a worked example’s length into the template; the shortness is the design, not a limitation of a
first pass. See lean-canvas_companion.md section 4 for the fuller argument; it is not repeated here.
One further note on scope, not on size: if more than one customer segment is genuinely in play, the originator’s own method is to sketch a separate canvas per segment rather than to widen one canvas to hold both. “Sketch a Lean Canvas for each customer segment”, starting with “the top 2-3 customer segments you feel you understand the best or find most promising”, then “lay them side by side and select the best business model to start with.” Treat a single canvas, filled in with this template, as covering one segment at a time.
The rubric
Section titled “The rubric”Score each row 0, 1 or 2. Below 12 out of 18, hold the canvas back from the one other person you are supposed to share it with once it is done: at least one box is still asking that reader to react to a feature list, a multi-year forecast, or a belief nobody could ever prove wrong, rather than to a real current best guess.
| # | Criterion | 0 | 1 | 2 |
|---|---|---|---|---|
| 1 | Early adopter, not market | Customer Segments names only a broad market or demographic label, with no Early Adopters line that differs from it | An Early Adopters line exists, but it restates the broader segment in other words rather than narrowing it | The Early Adopters line names a trait specific enough to recognize that person before a mainstream buyer exists at all |
| 2 | Problem stated as problem | At least one line in Problem is already a solution in disguise; lifted into the Solution box, it would read the same way | Every line is a problem, but at least one is generic enough to fit almost any product in any category | Every problem is a pain the segment has today, independent of this product, that could still be checked as true even if this product never existed |
| 3 | Solution bound late | Solution covers ground the Problem box never named, reading like a specification rather than the simplest fix | Solution matches the scope of Problem, but nothing in it shows that Customer Segments and Problem were actually settled first | Every line in Solution answers one specific problem named above, and the box could not have been written before that problem was |
| 4 | Near-term, not a forecast | Revenue Streams or Cost Structure states a multi-year figure, a projection, or a budget for a team the product has not earned | A near-term figure appears in both boxes, but neither is tied to a stated, specific milestone | Revenue Streams and Cost Structure name the same specific milestone, three to six months out, and say whether the price has actually been tested with anyone |
| 5 | Key Metrics are outcomes | At least one metric counts something the team produced, such as a feature shipped or an item added to a database, rather than a result | Between three and five metrics are listed, but at least one still counts an output rather than an outcome | Three to five metrics are listed, and every one is a result a customer or the business actually experiences, not an output the team merely produced |
| 6 | Unfair Advantage is honest | The box names effort, enthusiasm, or a feature a well-funded competitor could ship next quarter | The box is left blank, with nothing said about what would have to become true before something qualified | The box either names something that would survive being said out loud to a competitor, or it is blank and states plainly what is still missing |
| 7 | One message | The box reads as a list of features or benefits rather than one sentence stating why the product is different | One message is stated, but the High-Level Concept line is missing, or is a slogan with no comparison in it | One sentence states the single reason to pay attention, and the High-Level Concept names a concrete “X for Y” comparison a stranger could repeat back unprompted |
| 8 | Riskiest belief is falsifiable | No row in Riskiest Beliefs and Next Tests names a test that could come back negative, or the evidence column names enthusiasm rather than evidence | At least one row names a real test, but the rows naming a customer or problem belief are not the ones treated as riskiest | The row treated as riskiest names a customer or problem belief, states honestly what evidence, if any, backs it, and names a next test specific enough that it could refute it |
| 9 | Versions record real change | A revision has happened but the Versions table is empty, or its one row says only that “the canvas was updated,” with no box or reason named | A row names which box changed, but not the belief the change followed from, or whether that belief was confirmed or refuted | At least one row names the box that changed, the belief from Riskiest Beliefs and Next Tests it followed, and whether that belief was confirmed or refuted |
Anti-patterns
Section titled “Anti-patterns”- Drafting the first sketch as a group exercise. The current teaching page names this directly and gives the fix in the same breath: “If you’re part of a team, avoid creating a Lean Canvas as a group exercise. Instead, have each team member create their snapshot first.” Reconcile snapshots afterward; do not start from a blank page with the whole team in the room.
- Skipping the canvas and keeping the beliefs in your head instead. The 2011 walkthrough names this as a real and common failure, not a hypothetical one: “A lot of entrepreneurs skip this step altogether which is NOT the solution.” The same text names the cost of skipping it: “Too many founders carry their hypotheses in their heads alone which makes it hard to systematically build and test a business.”
- The solution-first trap: writing the Solution box before Customer Segments and Problem are settled. The instruction to wait is stated in both eras of Maurya’s writing: “Bind a solution to your problem as late as possible.” A trade-press interview reports him naming the general habit directly, including himself among those who fall into it: “The biggest mistake I find which we all make - myself included - is this love for the solution,” which he calls “an innovator’s bias”.
- Treating Revenue Streams and Cost Structure as a multi-year forecast. The current teaching page names the fix in the same sentence as the problem: “Rather than thinking in terms of three- or five-year forecasts, it’s better to take a more stage-based approach.” Both boxes get their near-term number from the same milestone, not from two separate timelines.
- Filling Unfair Advantage with effort or enthusiasm instead of an honest blank. “A real unfair advantage is something that cannot be easily copied or bought,” in Jason Cohen’s definition, which Maurya quotes. It is normal for the box to have nothing yet: “Leave the unfair advantage blank.” Writing in something that is really just hard work or enthusiasm is worse than leaving the box honestly empty.
- Marking a belief “validated” on the strength of excitement rather than a real test. Maurya’s own teaching on this practice is to “focus on your weakest links first - your riskiest assumptions”, and to check first whether customer and problem beliefs are grounded in real learning before trusting the rest of the model. A row that cannot be proven wrong, backed only by how excited the team already is, has not done that work.
- Sketching it once and never returning to add a Versions row. The canvas is built to be kept current, not archived after one pass: “The canvas is meant to be an organic document that evolves over time”, a format meant for “tracking ongoing learning”. A Versions table with no entry after the first sketch is a sign this discipline has lapsed, not a sign the model turned out to be settled on the first try.
Before you share it
Section titled “Before you share it”- You sketched this alone, or each team member sketched a snapshot first, before anyone reconciled them into one canvas.
- Every line in Problem states a pain the segment has today, not a solution wearing a problem’s heading.
- Revenue Streams and Cost Structure both name the same near-term milestone, not a multi-year number.
- Key Metrics lists three to five outcomes, not outputs the team produced.
- Unfair Advantage either names something a competitor could not simply copy or buy, or is honestly blank.
- The riskiest row in Riskiest Beliefs and Next Tests names a test that could actually come back negative.
- You have deleted every HTML comment before sharing it.
The artifacts
Section titled “The artifacts”lean-canvas_template-lean.md · ~6,250 tokens
---title: "{{product_name}} Lean Canvas"product: "{{product_name}}"owner: "{{who_sketched_this_canvas}}"status: "{{draft_or_tested}}"last_updated: "{{date}}"doc_type: lean-canvassize: leansource_template: lean-canvassource_template_version: 0.1.0---
<!--LEAN CANVAS. This bundle ships one size, deliberately: Ash Maurya publishes one page and says"Embrace a 1-page constraint", and no source this research read splits a short canvas from a longone. See lean-canvas_guide.md, "Picking a variant," for the fuller argument; it is not repeatedhere.
WHAT THIS DOCUMENT IS, AND IS NOT. A Lean Canvas states a product's business model on one page, asbeliefs to test, not as a finished plan. Ash Maurya built it as "an adaptation of AlexanderOsterwalder's Business Model Canvas - optimized for the Lean Startup methodology", keepingOsterwalder's layout on purpose to credit the source. It is NOT a business plan: "While a businessplan can take weeks or months to write, your initial canvas should be sketched quickly." It is NOTa finished forecast of Revenue Streams or Cost Structure; both boxes below ask for a near-termmilestone, not a multi-year model. It is NOT a product-strategy document: this canvas names nodecision about which problems to pursue, only the whole business model as a set of beliefs. It isNOT the Business Model Canvas, the tool it adapts: that tool keeps Key Partnerships, Key Resources,Key Activities and Customer Relationships, which this canvas replaces with Problem, Solution, KeyMetrics and Unfair Advantage. See lean-canvas_companion.md section 1 and section 8 for the fullerset of boundaries, including this library's own argued boundaries with `product-strategy`,`positioning-statement`, `north-star-metrics` and `lean-ux-canvas`, none of which any source readstates directly.
WHO FILLS THIS IN, AND HOW FAST. Sketch the first canvas alone, even on a team: "If you're part ofa team, avoid creating a Lean Canvas as a group exercise. Instead, have each team member createtheir snapshot first." Then share it: "The important thing is sharing it with at least one otherperson when you are done." A first sketch takes "less than 20 minutes", in "one sitting", on onepage, and it is normal to leave boxes blank: "it's okay to leave boxes blank". Maurya built thecanvas "for startups operating under conditions of extreme uncertainty", and his own company drawsthe line at product-market fit, not company age: "If you are pre-product-market fit, start withLean Canvas". See lean-canvas_companion.md section 6 (Debates) and section 9 (Adaptations) beforeusing this for an established, proven line of business.
WHERE THIS SITS IN THE CHAIN. This library places the Lean Canvas second among its strategy-docsfamily, between the product vision and the product strategy, and names that placement its ownPOSITION rather than a sourced finding: no source compares the two sequences by name. A consideredalternative places the canvas after the strategy, as the strategy's business model kept currentbeside the roadmap. Nothing in this template depends on which place is correct; it depends only onthe canvas stating beliefs a later product strategy can commit to or reject. Seelean-canvas_companion.md section 6 (Debates) and section 8 (Relationships).
THE CANVAS IS COPYRIGHTED. No open licence appears on Maurya's own pages. This template names theboxes and quotes him briefly with attribution; it does not reproduce his visual grid layout.
A NAME COLLIDES INSIDE THIS LIBRARY. `product-vision_guide.md` uses the phrase "the lean canvas"for that bundle's own short, one-page `canvas` format, an unrelated meaning of the same two words.This document is Ash Maurya's Lean Canvas, described fully in lean-canvas_companion.md section 8.
HOW TO FILL THIS IN1. Read the comment under each heading: WHAT it wants, WHY it matters (with a pointer into lean-canvas_companion.md for the deep reasoning), guiding questions to ASK, a GOOD and a WEAK example, and the TRAP to avoid. The last two sections also carry PRIORITY and ROW HINT for their table.2. Replace each {{placeholder}} with your content. Maurya's own fill order starts from Customer Segments and Problem, together, and this template follows that order top to bottom; he also says "There is no right order for sketching a Lean Canvas", so treat the order as a default, not a rule.3. If a section does not apply, write "N/A" and one line of why, rather than deleting it.4. This is a snapshot of your current best guess, not a finished plan: "Like everything on the canvas, you start with the best guess and iterate from there." Before you first share it, self-grade against lean-canvas_guide.md, then DELETE every HTML comment. They are guidance, not content.-->
# {{product_name}} Lean Canvas
## Customer Segments
<!-- WHAT Who has the problem and who will pay for a solution to it, with a line for the Early Adopters, the specific subset you target first. WHY Maurya's product page defines both directly: "Customer Segments - Who has the problem and who will pay" and "Early Adopters - The specific subset of customers you target first". His current teaching page draws the paying customer apart from a mere user: "A customer is someone who pays for your product. A user does not." Both his current teaching page and his 2011 walkthrough start the fill order from this box, and the walkthrough ties it to the next box directly: "The "Problem-Customer" pair usually drives the rest of the canvas which is why I tackle them together". Deep dive: lean-canvas_companion.md section 3 (Customer Segments). ASK Who exactly has this problem, in concrete terms rather than a demographic label? Who will actually pay for a solution, as distinct from who will merely use it for free? Within that segment, who is the early adopter: the person who will adopt before anyone else does, not the eventual mainstream buyer? GOOD "Busy home cooks who already use a grocery delivery app but still spend 20 minutes every evening deciding what to cook. Early Adopters: cooks who have already tried and abandoned at least one other meal-planning app in the past year." (names who pays, and narrows to a specific early adopter rather than the whole market) WEAK "People who cook dinner." (no payer named, and no early adopter named apart from the mainstream market) TRAP Writing the eventual mainstream buyer here instead of the early adopter. Maurya is direct about this: "Your objective is to define an early adopter, not a mainstream customer." -->
{{customer_segments}}
**Early Adopters.** The specific subset of customers you target first: {{early_adopters}}
## Problem
<!-- WHAT The top one to three problems this segment has, with a line for Existing Alternatives: how these customers solve the problem today, without your product. WHY "List the top one to three problems." Maurya's reasoning for keeping the box narrow is explicit: "Problems, not solutions, create spaces for innovation." The Existing Alternatives sub-block asks a separate question: "Document how you think your early adopters currently address these problems today." An alternative does not have to be a named competitor or even another product; doing nothing at all can be an alternative too. Deep dive: lean-canvas_companion.md section 3 (Problem). ASK What are the top one to three problems this segment actually has, stated as problems, not as the solution you already want to build? How does this segment solve each problem today, including an alternative that is not a competing product at all? GOOD "1. Deciding what to cook takes longer than actually cooking it. 2. The planned meal often does not match what is already in the fridge. 3. The grocery list gets written from a recipe, not from what is already on hand. Existing Alternatives: a paper list on the fridge door, a group text to whoever is cooking tonight, or skipping the plan entirely and ordering takeout." (names problems, not features, and gives a real existing alternative, including doing nothing) WEAK "People need a better way to plan meals." (names the shape of a solution, not a problem, and names no existing alternative) TRAP Naming a solution here instead of a problem. If the "problem" you wrote is already the answer you want to build, the rest of the canvas has nothing genuine left to test it against; this is the same solution-first trap the Solution box below is built to resist. Draft this box together with Customer Segments, not after it; the pairing is deliberate. -->
{{top_problems}}
**Existing Alternatives.** How customers solve these problems today: {{existing_alternatives}}
## Unique Value Proposition
<!-- WHAT The single message that states why this product is different and worth attention, with a line for the High-Level Concept: an "X for Y" analogy that makes the idea instantly understandable. WHY Maurya's product page defines the sub-block directly: "High-Level Concept - The X for Y analogy that makes your idea instantly understandable". His 2011 walkthrough places the Unique Value Proposition itself at the centre of the canvas and treats it as one of the hardest boxes to get right, though no line of his on that point survives as a direct quotation, so treat any stronger claim about this box with caution. Deep dive: lean-canvas_companion.md section 3 (Unique Value Proposition). ASK Why is this product different from every existing alternative named above, in one clear message rather than a feature list? What single analogy would make a stranger understand the idea in one sentence? GOOD "Hearth is the app that turns what is already in your fridge into tonight's dinner, in under a minute. High-Level Concept: a meal planner for people who already shop, not for people who want to shop differently." (one clear message, plus a concrete analogy a stranger could repeat back) WEAK "Hearth helps you eat better." (no stated reason to pay attention, and no analogy) TRAP Writing a feature list instead of one message. If this section needs more than a sentence or two, the message is not yet found; keep working it rather than padding the box with everything the product does. -->
{{unique_value_proposition}}
**High-Level Concept.** The X for Y analogy that makes the idea instantly understandable:{{high_level_concept}}
## Solution
<!-- WHAT The simplest thing that addresses each listed problem, in a box kept deliberately small. WHY "I purposely made the solution box less than one-ninth of the entire canvas". The instruction to wait is stated in both eras of Maurya's writing: "Bind a solution to your problem as late as possible." Deep dive: lean-canvas_companion.md section 3 (Solution) and section 6 (Debates, the solution-first trap). ASK What is the simplest version of a solution that addresses the problems above, not the most complete one? Have Customer Segments and Problem actually been settled first, or is this box being filled from a solution already decided on? GOOD "Photograph the fridge shelf; the app matches it against three dinner options already in the house." (one simple mechanism, stated small) WEAK A paragraph-long feature list covering onboarding, notifications, social sharing and a recipe marketplace, reading like a product specification rather than the simplest version of the idea. TRAP Filling this box before Customer Segments and Problem are settled. Writing a solution early is the specific habit this companion calls the solution-first trap; a founder who finds this box too small to hold a full solution description is being told something about scope, not about formatting. -->
{{solution}}
## Channels
<!-- WHAT The path to reach customers, now and at scale. WHY This is the thinnest-sourced box in the research behind this bundle. Maurya's 2011 text advises selling manually before automating, and selling in person before letting others sell on the founder's behalf, though no line of his on this point survives as a direct quotation, so it is reported here as paraphrase. Where this box is thin on texture, lean on the current teaching page's box order rather than inventing detail the research did not find sourced. Deep dive: lean-canvas_companion.md section 3 (Channels). ASK How will the early adopters named above actually be reached today, one at a time if necessary? What is the first channel that gets this in front of a real early adopter this week, not the channel that would work once the product is proven? GOOD "Early adopters reached one at a time through a cooking forum and a manual waitlist; no paid channel yet." (a specific, near-term, founder-reachable path) WEAK "Marketing." (names no actual path to a single customer) TRAP Writing a full go-to-market plan here, or a channel strategy that assumes scale the product has not earned yet. No source read names a channel framework that extends this box; keep it a short, founder-written line. -->
{{channels}}
## Revenue Streams
<!-- WHAT How the product makes money, priced now rather than modeled years out. WHY "Rather than thinking in terms of three- or five-year forecasts, it's better to take a more stage-based approach. Focus on your most immediate short-term milestones three to six months from now." Maurya's 2011 walkthrough is direct about timing: "Second, I believe pricing is one of the riskiest (and most critical) parts of the business model and should be tested early." Deep dive: lean-canvas_companion.md section 3 (Revenue Streams). ASK What is the near-term price, and the near-term milestone it is tested against, three to six months out? Has this price actually been tested with a real early adopter, or is it a guess that has not yet been put in front of anyone? GOOD "$9 a month, tested against the first 20 waitlist signups; next milestone is 10 paying households within three months." (a near-term price, tied to a near-term, testable milestone) WEAK "A freemium model, pricing to be decided later, with a five-year revenue projection attached." (no near-term price, and the exact multi-year forecast this box is built against) TRAP Writing a multi-year revenue model instead of a near-term price and milestone. This box shares its sourcing with Cost Structure below; the same stage-based instruction governs both. -->
{{revenue_streams}}
## Cost Structure
<!-- WHAT What it costs to reach the next milestone, three to six months out, not a multi-year budget. WHY The same stage-based instruction that governs Revenue Streams governs this box: "Focus on your most immediate short-term milestones three to six months from now." Like Channels, this box has thin sourcing beyond that one shared instruction; do not over-read structure into it that the source material does not carry. Deep dive: lean-canvas_companion.md section 3 (Cost Structure). ASK What does it actually cost, in money and in time, to reach the milestone named in Revenue Streams above? Is this a near-term figure, or has it quietly become a budget for a team and a timeline the product has not earned yet? GOOD "Fridge-matching API calls and two part-time cooks reviewing suggested meals, enough to reach 10 paying households by month three." (a near-term cost, tied to the same milestone as Revenue Streams) WEAK "A three-year operating budget with headcount for a 12-person team." (the exact multi-year budget this box is built against) TRAP Pairing this box with a revenue model that has already drifted into a multi-year forecast. Draft Revenue Streams and Cost Structure together against the same near-term milestone, not against separate timelines. -->
{{cost_structure}}
## Key Metrics
<!-- WHAT Three to five outcome metrics that show the business model is working. WHY "List 3-5 key metrics", and the current teaching states a clear preference: "Prefer outcome metrics versus output metrics." This box has changed name and content over time: the first published canvas headed it "Key Activity" and asked about a single key action rather than a set of outcome metrics. Fill it to the current instruction, not to that earlier wording. Deep dive: lean-canvas_companion.md section 2 (Origins and evolution) and section 3 (Key Metrics). ASK What three to five metrics would show, in outcomes rather than mere activity, that this business model is actually working? Does each metric measure a result for the customer or the business, rather than counting an output the team merely produced? GOOD "Weekly active meal plans created; percent of planned meals actually cooked; week-four retention." (outcomes a customer or the business actually experiences) WEAK "Number of recipes in the database." (an output the team produced, not an outcome that shows the model is working) TRAP Treating this box as a single key activity, the way the first published version of this canvas asked for. The current instruction is three to five outcome metrics, not one activity count. -->
{{key_metrics}}
## Unfair Advantage
<!-- WHAT Something about the business that cannot be easily copied or bought. WHY "A real unfair advantage is something that cannot be easily copied or bought" (Jason Cohen, as Maurya quotes him). The box's purpose is forward-looking rather than a pass or fail test: it exists "to continually encourage you to work towards finding/building your unfair advantage". It is normal to leave this box blank at first: "Leave the unfair advantage blank." Deep dive: lean-canvas_companion.md section 3 (Unfair Advantage). ASK Is there anything about this business, right now, that a well-funded competitor could not simply copy or buy? If nothing honestly qualifies yet, what would have to become true before something did? GOOD "Left blank for now; nothing yet stops a larger app from copying the fridge-photo matching feature." (honest, and names what would need to change) WEAK "We work harder than anyone else." (not something that cannot be copied or bought, and not an honest blank either) TRAP Filling this box with a claim that is really just effort, enthusiasm, or a feature a competitor could ship next quarter. A true unfair advantage survives being named out loud to a competitor; most early businesses have not found one yet, and leaving the box blank says so honestly. -->
{{unfair_advantage}}
## Riskiest Beliefs and Next Tests
<!-- WHAT A record of which beliefs on this canvas are weakest, what evidence, if any, currently backs them, and what test comes next. This section is this library's own addition to the nine boxes Maurya publishes: no source read publishes a canvas template with a tenth section of this kind. POSITION. WHY What this section records is sourced directly to Maurya's own teaching, layered on top of the nine boxes rather than inside them: "focus on your weakest links first - your riskiest assumptions", checking first whether "If your customer and problem assumptions are not yet grounded in empirical learning, it’s prudent first to get them in order." His product page frames the whole canvas around this discipline: "identifying and testing your riskiest assumptions before you build", and his company's own comparison page states it even more bluntly: "Find the riskiest assumption, test it". The prompt for what comes next is his own question: "what are the next hypotheses you need to test to move your product forward?" Deep dive: lean-canvas_companion.md section 3 (Riskiest Beliefs and Next Tests). ASK Of every belief written into the nine boxes above, which one would do the most damage if it turned out false? What evidence, if any, already backs that belief, as opposed to hope or enthusiasm? What is the next concrete test that would confirm or refute it? PRIORITY Name customer and problem beliefs first; Maurya's own advice is to get those grounded in empirical learning before trusting the rest of the model. A canvas with every belief marked "untested" is honest; a canvas that marks everything "validated" on day one is not. ROW HINT A good row names a belief from one specific box above, states plainly what evidence, if any, currently backs it, and names a next test that could actually refute it. A weak row names a belief nobody could prove wrong, or claims evidence that is really just enthusiasm from the people who already want to build this. GOOD | Problem | "The grocery list gets written from a recipe, not from what's in the house" | 20 waitlist signups said yes when asked; no usage data yet | Watch five households use a paper version of the matching for one week | WEAK | Problem | "People will love this" | Everyone we've talked to is excited | None named | (not falsifiable, the evidence is enthusiasm, and no next test is named) TRAP Marking a belief "validated" on the strength of excitement rather than a real test, or leaving the Next Test column blank for the riskiest row on the canvas. -->
| Box | Belief | Evidence So Far | Next Test ||---|---|---|---|| {{belief_box}} | {{belief}} | {{evidence_so_far}} | {{next_test}} |
## Versions
<!-- WHAT A dated row per revision: what changed on the canvas and why. This section is this library's own addition, chosen to match the `utility` classification this family contract gives the canvas: a standing instrument that earns its value by being kept current, not a document finished once. POSITION. WHY The practice it records is sourced: the canvas is "an organic document that evolves over time", built for "tracking ongoing learning", and Maurya's own software tool now automates exactly this: "Your canvas is versioned automatically. You can view past versions and restore any previous state." Deep dive: lean-canvas_companion.md section 3 (Versions). ASK What changed since the last recorded revision, and which box did it change? Did the change follow a test result named in Riskiest Beliefs and Next Tests above, or did it happen without one? PRIORITY List the most recent revision first, so a reader can see immediately whether this canvas has actually been kept current or has sat untouched since its first sketch. ROW HINT A good row names which box changed, ties the change to a belief that resolved, confirmed or refuted, and says so in one line. A weak row records that "the canvas was updated" without naming what changed or why. GOOD | 2026-03-01 | Narrowed Customer Segments to home cooks who already use a grocery delivery app | The first five interviews showed the mainstream buyer was not who actually signed up | WEAK | 2026-03-01 | Updated the canvas | Things changed | (names no box, no belief, and no reason) TRAP Sketching the canvas once and never adding a row after. A canvas with no Versions entries after its first sketch is a sign this discipline has lapsed, not a sign the model is settled. -->
| Date | What Changed | Why ||---|---|---|| {{version_date}} | {{what_changed}} | {{why_changed}} |---title: "Acme Embedded Analytics Lean Canvas"product: "Acme Embedded Analytics (exploratory, inside Acme)"owner: "Noor Siddiqui, Product Lead"status: "draft"last_updated: "2026-01-21"doc_type: lean-canvassize: leansource_template: lean-canvassource_template_version: 0.1.0---
> **Worked example.** A filled `lean-canvas`, the bundle's only size, for a new embedded-analytics OEM line> of business inside Acme, the company behind the `product-vision`, `product-strategy`,> `positioning-statement`, `north-star-metrics`, `prd`, `test-plan`, `test-case` and `bug-report` examples> elsewhere in this library. Per [ADR 0083](../../docs/internal/decisions/0083-lean-canvas-joins-strategy-docs-as-a-seventh-member.md),> this type sits second in the `strategy-docs` chain, between the vision and the strategy: the first sketch> below is dated 2026-01-19, five days after the> [product vision](../product-vision/product-vision_example.md) (2026-01-14), and the revision is dated> 2026-01-21, a week before the> [FY26 product strategy](../product-strategy/product-strategy_example.md) (2026-01-28). That strategy> names this same embedded-analytics OEM channel a non-target for Acme's core self-serve product, "both> requiring a different product than the one above." This canvas is a first sketch of that different> product, tested as its own business model rather than folded into the core strategy's scope.>> All figures are illustrative. They are internally consistent with the other examples in this library and> are not drawn from a real company.
# Acme Embedded Analytics Lean Canvas
## Customer Segments
Mid-market software vendors who already sell to the same operations, fulfilment and finance buyers Acme'sown core product serves, and whose customers have started asking them for charts and dashboards insidetheir product instead of a separate login somewhere else. These vendors do not want to be in the analyticsbusiness themselves; reporting is a feature their customers expect, not the product they set out to build.
**Early Adopters.** The specific subset of customers you target first: the two vendors our services teamalready built one-off embedded dashboards for in 2025, plus the four vendors on the core sales team'srunning list of deals that were marked "needs a different product" and routed away over the last twoquarters without ever being offered anything.
## Problem
1. Their own customers keep asking for dashboards inside the vendor's product, not a separate Acme login, and the vendor has no good answer for that request today.2. Building and maintaining a charting and data-modelling layer well enough to ship is a second product their engineering team did not sign up to build, and it competes for the same roadmap slots as their actual product.3. Every time one of these vendors scopes building it themselves, the estimate grows, because the first pass always undercounts what query performance and data freshness actually take to get right.
**Existing Alternatives.** How customers solve these problems today: bolt an open-source charting libraryonto their own database and hand-roll the queries, or tell their own customers to export a CSV and buildthe view somewhere else entirely. Both of the two vendors we built for in 2025 tried the open-source routefirst and stalled for more than six months before either one called Acme's services team.
## Unique Value Proposition
An analytics engine already proven against real customer data, available to drop inside another company'sown product under their own navigation and their own brand, rather than a from-scratch integration projecttheir engineering team has to staff and maintain.
**High-Level Concept.** The X for Y analogy that makes the idea instantly understandable: a paymentprocessor, but for analytics. The vendor keeps the login screen, the brand and the customer relationship;the dashboard rendering underneath it is ours.
## Solution
License API access to the same modelling and query engine the core product already runs in production,returned as embeddable chart components that the vendor's own front end renders inside their existingnavigation. Nothing new to build from zero: the two 2025 integrations become the first referenceimplementation, generalized into one API instead of two separate one-off builds.
## Channels
Direct outreach to the four vendors already sitting on the sales team's "needs a different product"referral list, plus one warm introduction from each of the two 2025 clients into their own vendor network.Nothing is bought or listed anywhere: every lead so far reaches Acme through a person who already knows it,which is how both 2025 deals began.
## Revenue Streams
A monthly platform fee plus a per-active-end-user usage charge, priced with the two 2025 deals' actualinvoiced total as a floor rather than a guess from zero. Milestone: one paying design partner signed on theusage-based contract, not the old flat project-fee arrangement, by the end of Q1 FY26.
## Cost Structure
One engineer pulled from the services team for the quarter to generalize the two existing one-offintegrations into a single reusable API, plus the hosting cost of running embedded dashboards for a smallnumber of design partners. Enough to reach the one-design-partner milestone above; not a budget for aseparate standing product team yet.
## Key Metrics
Design partners signed on the usage-based contract; weekly active end users inside each partner's embeddeddashboard; dollars billed on usage against the flat-fee floor named above; and the share of support ticketsthat land on Acme's own team rather than the vendor's, which tells us whether we are still selling a serviceinstead of shipping a product.
## Unfair Advantage
Empty, honestly. The two 2025 integrations show the engine can run inside another vendor's product, and anyembedded-analytics specialist could pitch the same four referral-list vendors tomorrow. Something wouldbelong here only once the overlap with Acme's own buyers proved real: if a vendor's customers already runAcme's core product, an embedded view could reuse the metric definitions they have agreed there, which anewcomer could not copy quickly.
## Riskiest Beliefs and Next Tests
| Box | Belief | Evidence So Far | Next Test ||---|---|---|---|| Customer Segments | A mid-market software vendor will pay an ongoing usage fee for an embedded dashboard rather than keep building their own charting layer | Two vendors paid a flat project fee for a one-off custom build in 2025; neither has yet been offered the usage-based version | Offer the usage-based contract to those same two vendors first, since they are the only ones who have already paid once || Problem | The real blocker for these vendors is engineering cost, not something else such as branding control or data residency | Both 2025 deals surfaced engineering cost in a single kickoff call each, not from a structured round of interviews | Run structured interviews with the four referral-list vendors before assuming the two prior deals generalize to the rest of the list || Revenue Streams | Usage-based pricing is viable at a number this market will actually pay, not just a number we can invoice | None yet. Both 2025 deals were flat fees negotiated case by case, so there is no usage-pricing data point at all | Price the first design-partner contract on usage and watch whether the pushback lands on the metering itself or only on the number |
## Versions
| Date | What Changed | Why ||---|---|---|| 2026-01-21 | Narrowed Customer Segments to mid-market vendors already selling to Acme's own operations and finance buyers, after sharing the first sketch with the services lead who ran both 2025 deals | The services lead pointed out that the two actual deals were never "any software company": both were vendors already selling into the same buyers Acme's core product already understands, and the broader market claim had no evidence behind it at all || 2026-01-19 | Sketched the first version alone. Customer Segments read "any software company wanting embedded analytics" | Written straight from the two 2025 deals, without checking whether a broader market exists beyond those two accounts |Provenance
Section titled “Provenance”The reasoning, the history and every source, in the repository:
- Companion - the long-form argument: why these sections, where the sources disagree, and what the bundle refuses to claim
- History - what changed in this bundle, and when
- Research log - every source consulted, with what each one actually supports
- Catalog metadata - the machine-readable record this page is generated from
Catalog record: 11 sections across 1 format(s), methodology Lean Startup, typically owned by Founder / PM.