What the nine blocks actually mean, the order to fill them in, and the mistakes that make most canvases decorative instead of useful — by the framework's own logic, not folklore.
The theory in one paragraph
A business model is how a company creates, delivers, and captures value — and the Business Model Canvas compresses that entire system onto a single page with nine linked blocks. Instead of a 40-page business plan nobody updates, you get a shared visual language: change one block and you can see which other blocks it strains. The canvas doesn't tell you whether your model works; it makes your assumptions visible enough to test — which is exactly why it became the starting artifact of the lean movement.
The mechanics — as Alexander Osterwalder defined them, not the folklore version.
Customer segments, value proposition, channels, relationships, and revenue — everything about who you serve and why they pay. This is the side that kills startups when it's wrong, so it gets filled in first.
Key activities, resources, partners, and costs — what it takes to deliver the right side. Founders love starting here because it's about their product. Resist that: the machine only matters once the market side holds.
The canvas earns its keep when you trace dependencies: a premium segment implies high-touch relationships, which implies expensive activities, which implies your cost structure. A canvas where the blocks don't reference each other is a sticky-note collage, not a model.
The canvas is a hypothesis board, not a summary
Osterwalder's own instruction is the one most founders skip: every entry on the canvas is a guess until evidence says otherwise. The teardowns in our benchmarks library work because the companies validated block after block — Uber's contractor model, Airbnb's trust machinery. Fill yours in pencil, then go test the block that scares you most.
Swiss business theorist · co-founder of Strategyzer
Osterwalder developed the canvas in his 2004 PhD thesis on business model ontology, then turned it into Business Model Generation — a book co-created with 470 practitioners from 45 countries. It made business model design a discipline with a shared vocabulary rather than an instinct.
Lineage — Extended by the Value Proposition Canvas, which zooms into the customer–value fit →
Each step maps to a field in the Business Model Canvas tool — finishing the read means finishing the work.
Name a specific segment — 'SME retailers in Riyadh who still invoice on paper', not 'businesses'. Then write the value proposition as the progress that segment makes, not your feature list.
Canvas · Customer Segments + Value PropositionChannels, relationships, revenue streams — in that order. If you can't name the channel that reaches your segment affordably, you've found your riskiest block early. That's the canvas working.
Canvas · Channels, Relationships, RevenueOnly list what's genuinely key — the activities that create the value proposition, not everything a company does. Uber lists lobbying; most canvases forget the uncomfortable blocks.
Canvas · left sideThe bottom two blocks are the profit logic. If costs scale linearly with revenue, you've drawn a services firm — fine, but know it. If they don't, you've drawn leverage; now prove it.
Canvas · Cost StructureCircle the blocks that rest on the least evidence. Those are your next month of work — customer interviews, a landing-page test, a manual pilot. The canvas is done when it's being falsified, not when it's full.
The steps above are the Business Model Canvas tool's structure. Open it and work through them with your own startup — it's free, no card required.
Free tool · free account · no card required
Teardowns from our benchmarks library where this framework is doing real work.
Benchmark teardown
Uber
A two-sided canvas where a partner choice (contractor drivers) IS the model
Read the teardown
Benchmark teardown
Careem
The same marketplace canvas, rebuilt for MENA constraints — and a $3.1B exit
Read the teardown
Benchmark teardown
Netflix
Watch one canvas replace another: DVDs → streaming → ads tier
Read the teardown
It's a one-page template that describes how a business works using nine blocks: who you serve (customer segments), what you offer them (value proposition), how you reach them (channels), how you keep them (customer relationships), how you earn (revenue streams), plus the activities, resources, partners, and costs needed to deliver all of that. It replaces long business plans with a single testable picture.
Start on the right side with customer segments and value proposition, then channels, relationships, and revenue streams — the market side. Only then fill in key activities, resources, partners, and cost structure. Starting with the market side forces you to confront the riskiest assumptions first.
The Lean Canvas is Ash Maurya's adaptation for early-stage startups: it swaps partners, activities, resources, and relationships for problem, solution, key metrics, and unfair advantage. Use the Lean Canvas pre-product when the problem is still unproven; use the Business Model Canvas once you're designing the full operating model.
Alexander Osterwalder, a Swiss business theorist, developed it from his 2004 PhD work on business model ontology and published it in the 2010 book Business Model Generation, co-authored with Yves Pigneur. It has since become the most widely used business-model tool in the world.
Yes — because it was never a document format, it's a dependency map. AI-era startups still have segments, channels, costs, and revenue logic; the canvas makes the links between them visible and testable. What has changed is expectations: a canvas today is expected to be validated block by block, not just filled in.
Lean Canvas
Ash Maurya · 2010
Value Proposition Canvas
Alexander Osterwalder & Yves Pigneur · 2014
The Lean Startup
Eric Ries · 2011
Sources
Independent educational summary written by StartupKit from public sources. Business Model Canvas is the work of Alexander Osterwalder; this page is not affiliated with or endorsed by the author.