The Lean Startup: Turn an Assumption into a Useful Test
TL;DR
In this guide
- Name the uncertainty. Write down what needs to be true for the idea to work.
- Choose the smallest useful test. Test the question without building unnecessary features.
- Use the result. Record what changed and decide what to do next.
The Lean Startup approach asks a team to learn about an uncertain business idea through tests rather than treating the first plan as proven. Start by deciding which assumption needs evidence.
The guide at a glance
The Lean Startup at a glance
Use these three ideas to guide the practical steps below.
01
Build
Create a limited test that can expose the assumption to real users or customers.
02
Measure
Collect evidence relevant to the question rather than only signs of activity.
03
Learn
Use the evidence to keep, change or stop a part of the plan.
What the method is trying to establish
Eric Ries describes validated learning and the build–measure–learn loop in the official Lean Startup methodology. The central question is whether the work teaches the team something useful about building a sustainable business.
The approach can be discussed within a new venture or an established organisation. It does not mean every business must launch quickly at any cost, or that a small first version is automatically a good test. Define the uncertainty before deciding what to build.
A minimum viable test needs a question
A minimum viable product should provide useful learning about a particular assumption. Listing fewer features is not enough. The team needs to know who will use the test, what behaviour or feedback matters and how the evidence will influence the next decision.
If the question is whether customers understand an offer, a prototype or short explanation may be suitable. If the question is whether an operational service can be delivered reliably, a visual mock-up alone will not answer it. Safety, privacy and contractual obligations still apply to a limited test.
An illustrative service enquiry experiment
Illustrative example: a business is considering a maintenance service. The assumption is that prospective customers understand the difference between a one-off repair and ongoing care. The team prepares a simple service description and asks relevant prospects to explain what they think is included.
Record the explanations, the questions raised and any mistaken expectations. Revise the description if the same ambiguity appears repeatedly. This test informs the offer wording; it does not establish market size, willingness to pay or long-term profitability.
If a live enquiry page is part of the test, confirm its form and response process before using it. A website brief helps distinguish the page’s job from the larger business assumptions.
Keep an experiment record
| Assumption | Evidence to collect | Decision after review |
|---|---|---|
| Customers understand the service | Their own explanation and questions | Clarify the offer or test a different description |
| The enquiry route is usable | Observed completion and failures | Fix the journey before drawing demand conclusions |
| The service can be delivered | A separately scoped operational trial | Revise the process or limit the offer |
These are illustrative records, not fixed success thresholds. Agree what would be useful evidence before looking at results. A small or unrepresentative test has limits.
Connect learning to the next investment
Innovation accounting makes progress visible through goals and milestones. Use a record of assumptions, tests and decisions so activity is not confused with learning.
Avoid repeating a test simply because it produced an encouraging number. Ask which decision is now better supported and what remains unknown. For choosing an appropriate next action, read how to use outside advice and build momentum.
Put it into practice
Before running a test
- Write one assumption in plain language.
- Identify the people and situation the test needs to represent.
- Choose evidence that could change the decision.
- Build only what is needed to collect that evidence responsibly.
- Record the limits of the test and the result.
- Agree the next action, owner and review point.
Keep in mind
A test is useful when it changes a decision
Visits, likes or a polished prototype can be encouraging, but the question is what they establish about the assumption you set out to examine.
Mistakes to avoid
- Building a small product without naming the uncertainty.
- Using convenient participants as though they represent every customer.
- Changing the success measure after seeing the result.
- Treating a limited test as proof of long-term business success.
