Chat with Jurre

Blueprints

Validation Canvas

Turn an assumption into a tested conclusion.

The Validation Canvas is a structured log for a single hypothesis: before you start, you record what 'success' means, and afterwards you decide explicitly whether to persevere, pivot or stop.

Fill in this canvas in BoldcasterOr start with the scale-up audit →
BoldcasterValidation Canvas · template

What is the Validation Canvas?

The Validation Canvas has its roots in Eric Ries's Lean Startup movement and Steve Blank's Customer Development method. Strategyzer, Mural and other training organisations have developed variants. The core stays the same: a testable hypothesis, the burden of proof needed to confirm or refute it, and the decision that follows.

The problem it solves: teams claim 'we have validated this' after hearing three customers nod politely in a conversation. Or worse, they 'validate' after the fact what they have already built. The canvas forces you to define in advance what counts as enough evidence, in what form, and which decision follows, so that 'validated' actually means something.

Where the Validation Canvas falls short: it is not an experiment design in itself. It structures the thinking around it (hypothesis, criteria, decision) but does not say which kind of experiment you should run. Combine it with an Experiment Canvas for each individual test, and with a Design Thinking process if you still know too little to formulate a hypothesis.

When do you use it?

Reach for the Validation Canvas as soon as you have a strong but untested belief: that this segment will pay, that this channel works, that this price will hold. A good rule of thumb: when someone on the team starts with 'I reckon…' or 'I think customers…', put it on the table as a hypothesis and bring out this canvas.

If you do not even have a hypothesis yet, start with a Design Thinking Canvas. Need to design a test for each hypothesis? Then the Experiment Canvas fits alongside it. Is it about a product feature? Combine it with the Feature Canvas.

How to fill it in

Work in order: first the hypothesis, then the critical assumptions, then the criteria, and only then the experiment. Write the decision only once you have the result. Do not change the order, or you undermine the whole canvas.

01

Hypothesis

A hypothesis is one testable statement in the form 'We believe that [audience] [shows this behaviour / has this problem / recognises this value] because [assumption]'. No opinions, no vagueness: a claim that can be factually confirmed or refuted. A good hypothesis specifies who, what and why; otherwise you cannot design the experiment that would break it.

Questions that help

  • ·What do we believe, and about whom exactly?
  • ·Which behaviour or outcome would support the hypothesis?
  • ·Can this be tested within four weeks?

Example

We believe that job planners at SME construction firms are willing to pay €99 a month for a planning tool with clash detection, because they currently lose two hours every week to rework caused by double bookings.

Common mistake

A hypothesis that is really a wish ('We believe this will be a great success'). It cannot be tested, so it is no use.

02

Riskiest Assumptions

Every hypothesis hides several assumptions. The riskiest is the one that brings the whole hypothesis down if it is wrong, and that you know least about. Say your hypothesis assumes that (a) there is a problem, (b) your solution solves it, (c) people want to pay for it, (d) you can reach them. Which is the most fragile? That is what you test.

Questions that help

  • ·Which assumption is the most fragile?
  • ·Which one do we have the least evidence for?
  • ·What would happen if this assumption turned out to be wrong?

Example

For the planning tool: the assumption that job planners make the buying decision themselves (rather than the owner) is the riskiest. If it falls away, your whole go-to-market is wrong.

Common mistake

Choosing a 'safe' assumption to test because it is easy to confirm. Test what would hurt you most if it were wrong.

03

Success Criteria

In advance, not afterwards, you record which result confirms the hypothesis. A concrete threshold: '7 out of 10 test customers choose the paid subscription within 14 days.' Not 'positive reactions'. Not 'we feel it is working'. Without a criterion fixed in advance, every result stays open to interpretation, and interpretation is usually wishful thinking.

Questions that help

  • ·Which number or behaviour proves the hypothesis is right?
  • ·What is an acceptable result, and what would be a disappointment?
  • ·Who signs off on this threshold in advance?

Example

Success criterion: at least 8 out of 15 visitors click 'buy now' after seeing the landing page; at least 3 of them actually complete the payment.

Common mistake

A vague criterion ('enough interest') or a criterion that gets adjusted afterwards. Both make the whole experiment meaningless.

04

Experiment

Here you describe how you will test it. Which type of test (landing page, interview, concierge MVP, price A/B test, prototype), which audience (how selected, how many people), which duration, which variables you hold constant and which you let vary. An experiment you cannot reproduce is not an experiment; it is an anecdote.

Questions that help

  • ·What is the cheapest way to test this?
  • ·How many people do I need for statistically meaningful data?
  • ·How do I make sure the outcome is not caused by something else?

Example

Two weeks of A/B testing a landing page through LinkedIn ads, targeting job planners at SME construction firms, budget €500, an expected 250 visitors per variant, main metric: conversion to 'book a demo'.

Common mistake

Designing an experiment that changes so many variables at once that afterwards you do not know what made the difference.

05

Result

Fill in this block only after the experiment has finished. Write down factually what happened: numbers, percentages, quotes, behaviour. No interpretation yet, only raw observations. Add context (duration, actual sample size, odd events during the test) so you can judge later how reliable the result is.

Questions that help

  • ·What was actually measured, without interpretation?
  • ·How large was the final sample?
  • ·Which external factors could have distorted the picture?

Example

Result: 247 visitors, 31 demo requests (12.5%), 7 payments after a demo (€693 in revenue). The threshold was 8 payments, so just below.

Common mistake

Mixing result and interpretation. 'It worked well' is not a result. Facts first.

06

Decision

The decision forces you and your team to choose explicitly. Three options: persevere (the assumption holds well enough, carry on), pivot (the assumption partly holds, adjust course) or stop (the assumption has been refuted, abandon this direction). One sentence, the date, and who made the decision. This prevents the most persistent startup disease: carrying on with a hypothesis that has in fact been refuted.

Questions that help

  • ·Based on the result: persevere, pivot or stop?
  • ·What is the first follow-up action within 7 days?
  • ·Who signs off on it, and who needs to be informed?

Example

Decision (12 March): pivot. We will target owners of construction firms as buyers instead of job planners. Next experiment: outreach through the networks of Bouwend Nederland, the Dutch construction trade association.

Common mistake

Taking a 'soft' decision ('we will keep an eye on it for a while'). The canvas loses its value if no clear go/no-go comes out of it.

Common mistakes

Formulating a hypothesis that is not a hypothesis

'Our customers love us' is not a hypothesis; it cannot be falsified. A hypothesis specifies who, what and how you would measure it.

Setting success criteria afterwards

The whole point of fixing criteria in advance is that you cannot hide behind interpretation. Do it afterwards and every result is 'actually quite positive'.

Too small a sample, too big a conclusion

Three pleasant conversations are not validation. Work out in advance how much data you need to say something meaningful; for behavioural metrics, ideally more than 100.

Avoiding the riskiest assumption

It is tempting to test the assumption you already believe first. But test what would hurt you most if it were wrong; that is where you learn the most.

Using a single channel or team as the audience

If you experiment within your own customer base or through one channel, you do not know whether it also works outside that segment. Always keep that in mind in the decision.

Sticking a 'validated' label on something that is no more than a signal

One positive test proves nothing definitive. The canvas takes you to 'enough confidence to take the next step', not to 'proven'.

In Boldcaster

How Boldcaster helps you with this

Fill it in interactively

No stray PDF lost in your downloads folder. Fill in the canvas in the app, with version history and comment threads.

Bold AI does the groundwork

Paste in your business plan, annual report or rough notes and Bold AI fills in the canvas for you. Or ask it for suggestions block by block.

Connected to your strategy

Send an insight from this Validation to your Priorities, KPIs or strategy in one click, instead of leaving it on a whiteboard.

All canvases in one place

Pilot users get access to BMC, Lean Canvas, Value Proposition Canvas and whatever comes next. One platform, one system.

Create a free account and use every canvas

Prefer some guidance?

You don't need to know the models: Bold AI helps you block by block. Want someone alongside you to run the sessions with your leadership team? Start with the scale-up audit. After that, we're happy to introduce you to a coach from our network who works with Boldcaster.

Frequently asked questions

What is the difference between a Validation Canvas and an Experiment Canvas?

+

The Validation Canvas zooms out to one hypothesis and the decision that follows from it. The Experiment Canvas zooms in on one concrete experiment within that hypothesis; you often need several experiments to test one hypothesis properly.

How many hypotheses should I test per quarter?

+

That depends on your stage. Teams before product-market fit ideally test a hypothesis every 1 to 2 weeks; mature organisations run heavier experiments less often, but with more rigour.

How do I write a good hypothesis?

+

Use the structure 'We believe that [who] [what] because [why]'. Make it specific enough that an outsider could judge after an experiment whether the hypothesis holds.

What if the outcome is unclear, just below or just above the criterion?

+

That is an important decision: run another test with a larger sample, or pivot to an adjacent hypothesis. Do not explain it away or interpret it optimistically.

Does every assumption need to be tested in a formal experiment?

+

No. Some assumptions are sufficiently supported by existing data, customer conversations or market reports. The canvas does make you name explicitly which evidence you have and which you still lack.

Who should be involved in writing a Validation Canvas?

+

At least someone with domain knowledge and someone with experience of experimentation. A product manager plus a designer or data engineer is a good combination.

How do I avoid confirmation bias when interpreting results?

+

First write down only the raw facts. Discuss what they mean in a second session, preferably with someone who is not emotionally invested in the hypothesis.

How long does a typical cycle take?

+

You fill in the canvas itself in 60 to 90 minutes. Running the experiment takes anything from a few days to several weeks, depending on the type of test and the sample size. Making the decision takes an hour, provided the criteria were clear in advance.

Further reading

What does a filled-in Validation Canvas look like?

An empty canvas is the easy part. The hard part is what goes in it. I'll send you a completed Validation Canvas: a fictional company, real wording, sharp enough to base decisions on.

See a completed example

The strategy canvas of a fictional installation company, fully completed. Plus five signs that a canvas actually steers the business, rather than just being filled in. Four pages, PDF.

You'll see the example straight away and get a copy by email. We keep your address so we can help with your strategy; unsubscribe any time in one click.