What each of the nine blocks means
Problem: the top one to three problems worth solving, plus how those people cope today — the existing workaround is your real competitor, not the other startups. Customer Segments: who has the problem, and separately, which early adopters feel it hardest right now. Unique Value Proposition: one sentence on why you're different and worth attention, written for a customer rather than an investor. Solution: the smallest thing you could build for each problem, one line each. The block is deliberately small, because at this stage the solution is the part you should be least attached to.
Channels: the routes to your customers, favouring ones you can test this month over ones that only work at scale. Revenue Streams and Cost Structure: how you charge and what it costs to run, including your own time. Key Metrics: the few numbers that tell you the thing is working — the activity that signals real value, not the totals that only go up. Unfair Advantage: something that can't easily be copied or bought. Most people have nothing here at the start, and writing "passion" or "great team" is worse than leaving it blank.
The order to fill it in
Problem and Customer Segments together, first — they only make sense as a pair, and a problem with no named holder is a hobby. Then Unique Value Proposition, which is where the pair gets stress-tested: if you can't write it in one sentence, the segment is usually too broad. Solution comes fourth, once the problem is pinned down, and Channels fifth.
Then the economics — Revenue Streams and Cost Structure, which are quick and mostly reveal whether the numbers can ever work. Key Metrics eighth, and Unfair Advantage last, because it is the block most likely to be honestly empty. Maurya's own advice is to time-box the first pass to about fifteen or twenty minutes, leave gaps where you genuinely don't know, and treat the blanks as your research list.
Common mistakes
Filling every block on the first pass is the big one. A canvas with no blanks usually means you invented answers, and the blanks are the most useful thing on the page — they are exactly what to go and find out. The second is skipping the existing alternatives in the Problem block: if you don't know what people do today, you can't tell whether your solution is enough of an improvement to make anyone switch.
Also common: vanity metrics in Key Metrics (registered users, page views) instead of the activity that shows someone got value; a Unique Value Proposition that describes the category rather than the difference; and an Unfair Advantage listing things a competitor could buy next week. And the canvas is a set of hypotheses, not evidence — its job is to make your assumptions explicit enough to test, so plan on rewriting it after every batch of customer conversations rather than framing it.
When Lean Canvas beats the Business Model Canvas
Choose the Lean Canvas when the riskiest thing you have is an assumption about people: pre-launch, a new product line inside a larger company, or any point where you cannot yet prove the problem is real. Its four substituted blocks all point at that risk — Problem and Key Metrics are about evidence, Solution is kept small on purpose, and Unfair Advantage forces an honest answer about defensibility.
Choose the Business Model Canvas once there is an operating model to describe — real revenue, real partners, infrastructure with a cost. Its Key Partners, Key Activities, Key Resources and Customer Relationships blocks are the ones that matter when the question has shifted from "does anyone want this" to "how does this run and scale". Plenty of teams use both: Lean Canvas while validating, Business Model Canvas once the thing works. Both are one-page tools, and neither replaces a financial model.
A canvas you'll actually revisit
The Lean Canvas is meant to change. Save it to OmniCanvas and it arrives as nine editable blocks on an infinite canvas, so the next version can sit beside the interview notes and the experiment results that prompted it — and you can see what you believed three months ago instead of overwriting it.