The canvas is most valuable when it helps you decide what to learn next. The goal is to replace important guesses with evidence before spending large amounts of time or money.

Step 1: Mark your assumptions

Read every square and ask: How do we know this is true? Label each important statement as one of the following:

Even validated statements can stop being true when markets or customer behavior change.

Step 2: Find the most dangerous assumption

An assumption is risky when it has high impact if false and little supporting evidence. A simple ranking system helps:

Question Low risk High risk
If false, how badly does it hurt? Small adjustment The business model may fail
How much evidence do we have? Repeated real behavior A few opinions or guesses
How costly is it to test later? Fast and inexpensive Slow and expensive

Test the assumptions that could cause costly failure before building around them.

Step 3: Design a small experiment

A clear experiment has five parts:

  1. Claim: What exactly do you believe?
  2. Audience: Which customer segment are you testing?
  3. Method: What will people actually do—interview, reserve, purchase, renew?
  4. Threshold: What observation would make you continue or change direction?
  5. Deadline: When will you review the evidence?

Example experiment

Claim: Design teams will pay $18 to rent a kit for a week.

Audience: Teams currently preparing their next physical prototype.

Method: Offer a small number of real rental kits with a working checkout process.

Threshold: Choose a target for paid rentals before the test begins, and compare actual payments with that target.

Review: After two weeks, decide whether to keep the price, change the offer, or revisit the customer segment.

Step 4: Distinguish different kinds of evidence

A friendly comment like “That sounds interesting” is weak evidence of a paying market. A customer making a repeat purchase is stronger evidence. Qualitative interviews are excellent for discovering problems and understanding motivations; observed behavior and transactions are better for testing adoption and price. Neither type alone answers every question.

Step 5: Revise the canvas

Version your work: v1.0, v1.1, v2.0. Record what changed and why. A business idea that improves in response to evidence is stronger than an attractive first draft that nobody challenges.

Keep a simple experiment log

Date Assumption Test Observation Decision
Today Customers have problem A Five interviews Three real examples Interview more people
Next week Price B is acceptable Small paid pilot Actual purchases Keep / change / stop

Do not fill the log with guesses about how participants might respond; record what they actually said or did.

Let your canvas evolve

Update the nine-box board after each meaningful discovery. Print or export dated versions to preserve the history of your thinking.

Revise the one-page canvas →