Innovation Design

Business Model Canvas

The business model canvas is one page for partners, activities, resources, the value proposition, relationships, channels, segments, costs, and revenues.

The questionWho pays for this business, for what, and what does it cost to serve them?

Matrix · Alexander Osterwalder and Yves Pigneur, 2010

Customer side

Customer segments Who the model serves, split by need
Value propositions What each segment gets that it would choose
Channels How a segment hears, buys, and gets help
Customer relationships How customers are won, kept, and grown
Revenue streams What each segment pays, for what, and how

Infrastructure side

Key partnerships Suppliers and allies the model depends on
Key activities What must be done repeatedly and well
Key resources Assets the model cannot run without
Cost structure Main costs, fixed apart from volume-driven

Schematic list of the nine blocks, grouped as the customer side and the infrastructure side.

When to use it

A client has a strategy deck, but nobody can say on one page who pays, for what, and what it costs to serve them.
You are assessing a new venture, a spin-out, or an acquisition target and need a shared description before anyone argues about the numbers.
One company runs two or more models, such as a product business and a platform, and the team keeps mixing them up.
A competitor has changed how it earns money, and you want to compare its model with the client's one block at a time.

When to avoid it

You need to know whether the market is attractive or how rivals will respond. The canvas has no explicit place for competition, so use an industry analysis alongside it.
The question is about a single process, an organization chart, or a pricing table. The canvas is too coarse for any of those.
The team will fill it in once and treat the result as a finding. It records claims and does not test them. Move to Lean Startup experiments if the claims are unproven.
The company has several segments with different offers and the plan is to put them all on one sheet. The blocks will blur.

Overview

The canvas puts a business model on one page in nine blocks, so a team can see who the customers are, what they are offered, how it reaches them, what they pay, and what it costs to deliver. Consultants use it to design a new venture or to describe an existing business before changing it.

It describes a model and does not test one. Every note is a claim until someone has checked it against customers, and the sheet looks equally tidy whether the notes are facts or hopes. It also has no block for competitors, so a model can look sound on the page and still be weak in its market. Keep one canvas per model, date each version, and treat the redraw as the main output.

The nine blocks

Customer segments

The groups the model serves, split wherever needs, willingness to pay, or ways of reaching them differ. Start here. In a two-sided business the people who use the service and the people who pay for it are separate segments, and each gets its own row of notes.

Value propositions

What the business gives each segment that the segment would choose over its alternatives: a job done, a cost removed, a risk reduced, access to something. Write one per segment, in terms the customer would use.

Channels

How a segment hears about the offer, buys it, receives it, and gets help afterward. Include routes that run through partners, since they carry cost and take some control.

Customer relationships

How the business wins customers, keeps them, and grows what they spend. Self-service, a named account manager, and a user community cost very different amounts, so this block sets how much margin the model needs.

Revenue streams

What each segment pays, for what, and how: a sale, a subscription, a usage fee, a license, a commission, advertising. List streams separately because they usually have different prices and different margins.

Key resources

The assets the model cannot run without. People, intellectual property, plant, data, brand, and capital all count. Include only what the value proposition really depends on.

Key activities

The things the business must do repeatedly and well to deliver the offer, such as making a product, solving client problems, or running a platform.

Key partnerships

Suppliers and allies the model depends on instead of owning. Note why each one is there: to get scale, to share risk, or to reach a resource the business lacks.

Cost structure

The main costs that follow from the resources, activities, and partners above, with fixed costs separated from costs that rise with volume. If this block surprises the client, an earlier block is probably wrong.

What it is not

A business plan. It describes how one model works and contains no market sizing, competitor analysis, or forecast.
A test of the model. A completed canvas is a set of claims, and a tidy page is easy to mistake for evidence.
A picture of the whole company. A firm with several segments and offers needs one canvas per model.

The steps

Scope one model and one segment

Decide which business model you are drawing and who its first customer segment is. If the company runs several models, run the exercise once for each. Title and date the sheet before anyone writes on it.
Fill the customer side first

Work through segments, value propositions, channels, relationships, and revenue streams, in that order. Mark every note as fact, belief, or guess. Most teams find that revenue and relationships contain more guesses than they expected.
Fill the infrastructure side

For each value proposition, ask what resources, activities, and partners it requires, and what those cost. Write only what the proposition needs. Anything on the page that supports no segment is either misplaced or not part of this model.
Walk each segment through the page

Follow one segment from value proposition to channel to revenue, then check that revenue covers the costs that segment causes. Look for blocks with no link to any segment, and segments that no channel reaches.
Rank the assumptions and pick tests

Turn the guesses into statements that can be checked, rank them by the damage if they are wrong, and decide how to test the top few. Redraw the canvas, dated and versioned, when the evidence arrives.

In practice

The owner of a regional equipment distributor asked for help with pricing. Before touching a price, the consultant had the management team draw the business twice: one canvas for contractors who buy machines, one for facilities teams who rent them. On a single sheet the two had looked alike. Drawn separately, nothing matched. Contractors came through field sales and bought on price. Facilities teams came through a web form and cared about uptime. The service technicians, listed as a key resource on both, had always been costed against the sales side.

Once the cost block was split, the rental side looked worse than reported, and the pricing question became a different one: whether rentals needed their own relationship model and a service-level price. The team wrote down three things that would have to be true for that to work and named who would check each. The canvas did not answer the question. It produced the assumptions to go and test.

What you leave with

A dated canvas for each business model, with every note marked as fact, belief, or guess.
A ranked list of the assumptions underneath it, ordered by how much damage each would do if wrong.
A named test, an owner, and a date for the top two or three assumptions.
A block-by-block comparison of the current model with one alternative, or with a competitor's.

Failure points

Notes without evidence

The team fills the page from memory in an hour. The result is internally consistent, and consistency gets read as validity. Marking each note by how it is known is the cheapest guard.

One canvas, several businesses

Segments with different propositions, channels, and cost drivers share one sheet, so every block holds a vague average. Nothing on the page can be tested because nothing is specific.

A value proposition that is a feature list

The block fills with product descriptions instead of what the customer gets done. If that happens, work on the proposition separately with the Value Proposition Canvas before returning to this page.

The workshop as the finish

The page is photographed, the deck is updated, and no assumption is tested. A canvas without a test list is a more attractive version of the plan it was supposed to replace.

No view of the competition

The canvas can look coherent in a market where an incumbent can cut price or copy the channel. Add a competitor's canvas or an industry analysis, because the sheet itself will not raise the question.

The failures above are usually Leverage & IP: it was used once, and nothing was left that makes the next engagement easier.

Origin

Alexander Osterwalder's 2004 doctoral thesis, The Business Model Ontology, supervised by Yves Pigneur, proposed a reference model drawn from the business model literature. The nine building blocks that became the canvas followed, and Osterwalder and Pigneur published the one-page layout in Business Model Generation (Wiley, 2010), written with input from a large community of practitioners. Strategyzer, the company built around their tools, maintains the canvas today. Ash Maurya's Lean Canvas is a well-known adaptation for early-stage ventures.

Strategyzer states that the canvas is released under a Creative Commons Attribution-ShareAlike 3.0 license. You may use it in client work and adapt it, as long as you credit Strategyzer visibly, and Strategyzer recommends contacting it before building the canvas into software. The Value Proposition Canvas is covered by a stricter copyright notice, so do not assume the same terms apply. This note describes the nine blocks in its own words and does not reproduce the official layout.