Business Model Canvas
A one-page model of how a business creates, delivers, and captures value across nine linked blocks — used to see whether the whole thing actually holds together.
Nine boxes anyone can fill in. The method is not the filling in. It is discovering that block three contradicts block seven, on paper, before it does so in the market.
The visual
Nine blocks. One structure. The links between them are the method.
The canonical nine-block canvas, but built as a coherence engine rather than a form. The blocks are visibly linked — which blocks constrain which, and what changes when one block moves. One link is already under tension.
What it is
The canvas is not nine boxes to fill in. It is a coherence test disguised as a template.
The Business Model Canvas describes, on a single page, how a business creates, delivers, and captures value. It does this across nine blocks: who you serve (Customer Segments), what you offer them (Value Propositions), how you reach them (Channels), how you relate to them (Customer Relationships), and what you earn (Revenue Streams), together with what you need to do it (Key Resources), what you do (Key Activities), who helps (Key Partners), and what it costs (Cost Structure). Together, they are a complete account of a business model, compressed into something you can see all at once.
But the canvas is almost universally misunderstood as a form, and this is the thing worth being clear about: the method is not filling in the boxes. Anyone can fill in nine boxes plausibly, and doing so produces a satisfying feeling of completion that is entirely unearned. The method is coherence. The nine blocks are not independent — they are linked, and they constrain each other. Choose a premium, high-touch channel and your cost structure changes. Choose a new customer segment and your value proposition, your channels, and your customer relationships all move with it, whether you update them or not.
The canvas earns its keep at the moment you discover that two of your blocks contradict each other. A business model that is internally inconsistent does not work, and the whole point of putting it on one page is to make that inconsistency visible before the market makes it visible for you, expensively. Filling in the canvas is the beginning of the work. A completed canvas is a hypothesis.
Try it
Change one block. Watch another break. That is the method.
Click any block to see what it holds and what it depends on — no block stands alone. Use the scenario buttons to see how changing one block propagates, and how a premium channel change produces a visible break in the cost structure. The break is the payoff.
When to deploy it
Use it when you need to see the whole model at once — and check whether it holds.
Use it when
- →You need to check whether the parts of a business model are consistent with each other.
- →You are changing one part of a business — a new segment, a new channel, a new revenue model — and need to see what else must change with it.
- →Different people in the room hold different models of the business, and you need one shared, visible artifact to argue against.
- →You are evaluating a new venture and want to expose untested assumptions in an organised way.
Do not lean on it when
- ×You want depth on the customer and the offer specifically — that is the Value Proposition Canvas, which zooms into a single block and does it properly.
- ×You want to explore the full range of innovation interventions beyond the model itself — the canvas describes what IS; other methods help you decide what to CHANGE.
- ×Filling it in will be treated as the work. A completed canvas is a hypothesis; if the organisation mistakes it for a plan, the canvas has done harm.
- ×The business is genuinely unknown and unformed. The canvas is excellent at organising what you believe; it can make a pile of guesses look like a strategy.
The honest limit
The canvas describes; it does not validate. Every block is a claim, and the canvas has no opinion about whether any of them are true. It is also static — a snapshot with no time, no competition, and no dynamics. And because it is fast, tidy, and universally recognised, it produces a powerful and often unwarranted sense that the business has been figured out. The canvas is a beginning.
How it works
Six steps — the last one is the only one that matters.
Start with the customer segment, not the product
Who, precisely, is this for? Everything else in the canvas hangs off this block, and a vague segment produces a vague and unfalsifiable model. If you cannot name who is NOT the customer, you have not chosen a segment.
Work the value-creation side, then the delivery side
Fill the customer-facing blocks first (segments, value propositions, channels, relationships, revenue), because they define what the business must be able to do. Then work the delivery side (key resources, activities, partners, costs), which is what it takes to actually keep that promise.
Make the links explicit, not just the blocks
For each block, ask what it depends on and what depends on it. This is the step teams skip, and it is where the method's value lives. A canvas without stated dependencies is a list of nine opinions, not a model.
Stress-test it: change one block and see what breaks
This is the actual method. Change the segment and see whether the value proposition, channels, and relationships still make sense. Change the channel and see whether the cost structure works against the revenue streams. If nothing breaks when you change something significant, you have not connected the blocks.
Find the contradiction, and take it seriously
Somewhere, two blocks will not agree. This is the finding. Do not resolve it by softening the language; resolve it by changing the business model, or by acknowledging that you have a real problem to solve.
Name the hypotheses inside each block
Every block contains claims that could be false. Write them down as assumptions and hand the riskiest ones to Assumption Mapping and Concept Testing. The canvas organises your guesses; it does not test them.
Best practices
If nothing broke, you did not use the canvas.
When it goes well
The mistakes
Treating it as a form.
Nine boxes filled in plausibly, admired, and filed. The filling in is not the method; the coherence check is. If nothing broke, you did not use it.
Mistaking a completed canvas for a validated one.
It is a hypothesis, and a tidy one, which makes it more dangerous rather than less. Every block is a guess until tested.
Changing one block and not the others.
Teams routinely swap in a new customer segment and leave the value proposition, channels, and relationships untouched — at which point the canvas is describing a business that does not exist.
A vague segment.
"SMEs", "consumers", "enterprises". Vague segments make every downstream block unfalsifiable, and the canvas becomes unfalsifiable with them.
Ignoring the contradiction.
When two blocks do not agree, the temptation is to reword until they seem to. The contradiction is the single most valuable thing the canvas produced. Do not talk it away.
Logistics
Do it together. Sketch several. Expect the disagreement.
Time
A focused session of 2–3 hours to map and stress-test. Revisited whenever a significant block changes — not on a calendar cadence.
Who
Cross-functional team of 4–8. The most valuable moment is discovering that people who have worked together for years hold quietly different models of the business. A canvas produced by one person records one person’s assumptions.
Approach
Sketch several canvases — the current model, then two or three alternatives — and compare. Write the assumptions on the canvas itself. Present it as a model to be attacked, not a plan to be approved. Keep the versions.
AI and this method
AI fills all nine blocks in seconds, beautifully. A completed canvas was never the achievement — and now it is not even work.
Toggle between modes to see what changes when AI generates the canvas — and the one use where it is a genuinely valuable adversary.
In practice
An enterprise software company moves down-market. One block changes. Almost everything else breaks.
Tab A shows the structured canvas working — mapping the current model, changing one block, following the dependencies, and finding the contradiction that tells them what the move actually requires. Tab B shows what happens when the canvas is produced by AI instead.
Shared scenario
A company that sells software to large enterprises through a direct sales team wants to move down-market and serve small businesses. Everyone in the room agrees it is a growth opportunity. They use the Business Model Canvas to see what the move actually requires.
Both tabs map the same business. Only the method differs.
Draw the current canvas first
The team started by mapping the existing enterprise model: large enterprise customer segment, high-touch direct sales channel, dedicated account management relationships, large annual contracts as revenue streams, and a cost structure built around an expensive sales force. Coherent. It worked, because every block supported every other one. The model held together.
Change one block and follow the dependencies
They changed one block: customer segment, from enterprise to small business. And then they did the thing the method is actually for, which is following the dependency links to see what moved with it. Almost everything did. The value proposition had to change — small businesses do not need, and will not pay for, the enterprise feature set. The channel had to change — a direct sales team cannot economically call on thousands of small accounts. The customer relationship had to change — dedicated account managers are impossible at small-business price points. The revenue stream had to change — large annual contracts become small monthly subscriptions.
⚠ The break — this is the method
With small monthly subscriptions as the revenue stream, the existing cost structure — built around an expensive direct sales force — simply did not work. The cost of acquiring a small-business customer through direct sales exceeded what that customer would ever pay. Two blocks contradicted each other outright.
That contradiction was the finding. It arrived on a whiteboard rather than in the P&L eighteen months later. The move down-market was not impossible — but it was not an extension of the current business model. It required a genuinely different one: self-serve channel, product-led onboarding, a fundamentally different cost structure. Now the team could argue about that, honestly, instead of about whether small businesses were an attractive market.
Write the assumptions on the canvas itself
They wrote the untested claims onto the new canvas: that small businesses would adopt self-serve, that the product could be simplified without losing what customers valued, that the freemium conversion rate would hold. Those went to concept testing and assumption mapping. The canvas organised the guesses. It did not settle them. A completed canvas is a hypothesis, and they knew that — which is why the work was not done when they finished filling it in.
Where it fits in the frameworks
The canvas appears at the moments where frameworks decide what the business will actually be.
The canvas holds the hypotheses the Build-Measure-Learn loop exists to test. Every block is an assumption — which segment, which channel, which revenue mechanism, which cost model. The canvas makes those assumptions explicit and organised before the loop begins, so each experiment targets a named claim rather than a vague belief.
In the Define phase, the canvas maps the business model behind the concept and checks whether it holds together. In Deliver, it is revisited as the model becomes concrete — ensuring that operational commitments (channel, cost structure, partnerships) are consistent with the value proposition and segment choices made earlier.
During concept development, the canvas is the business model behind the concept, made explicit and checkable. At the business case stage, the coherence between blocks determines whether the concept is investable — not just desirable, but financially viable as a whole model.
Design Thinking addresses desirability, feasibility, and viability. The canvas is the viability half of that question — does the business model that delivers the desirable solution actually work? It is the structured check that the viable business model exists, not just the desirable product.
Related methods
Methods that work alongside, before, or after the canvas.
The nesting sibling, by the same authors. The Value Proposition Canvas zooms into one block — the value proposition against the customer segment — and does it in depth, forcing an honest match between what customers need and what you offer. The Business Model Canvas is the whole board, and its concern is whether all nine blocks fit together. Use the VPC to get one block right; use the BMC to see whether the whole thing stands up.
The essential downstream partner. Every block on a canvas contains untested claims, and a completed canvas is a well-organised set of guesses. Assumption Mapping is how you sort those guesses by importance and uncertainty and decide which to test first. The canvas organises the hypotheses; Assumption Mapping prioritises them.
Where the canvas's riskiest assumptions go to meet reality. Coherence is not truth, and a perfectly consistent business model can still be one nobody wants. Concept Testing checks whether the value proposition, the channel, and the revenue model actually work with real customers — not just with each other.
Complementary lenses on the same object. The 10 Types framework diagnoses where innovation is happening — or not happening — across ten dimensions from configuration to experience. The canvas describes how the business model works as a whole. Together they show both what the model is and where it is and is not being innovated.
Sources and further reading
The books behind the method.
Osterwalder, A., & Pigneur, Y. (2010). Business Model Generation. Wiley.
The book that introduced the canvas, and still the clearest guide to using it. The nine-block structure, the canonical arrangement, and the visual grammar are all here.
Osterwalder, A., Pigneur, Y., Bernarda, G., & Smith, A. (2014). Value Proposition Design. Wiley.
The companion volume that goes deep on the single most important block — the value proposition against the customer segment — and provides the Value Proposition Canvas as the tool for that depth.
Bland, D., & Osterwalder, A. (2019). Testing Business Ideas. Wiley.
The essential counterpart: how to test the assumptions a canvas exposes, rather than admiring the canvas. Turns the hypothesis list a completed canvas produces into a structured testing programme.
Strategy & Prioritization — Method 9 of 9
40 methods across 6 stage groups