What a product discovery sprint includes, and what you get at the end
What a product discovery sprint should decide, what you need to bring to it, and the handover you should own at the end, written for a founder without a product team.
A product discovery sprint is a short, fixed piece of work that tests a product idea before anyone builds it. It should end with a decision to build, change or stop, and a set of documents you own: a problem statement, research notes, a scoped first version, a clickable prototype, a build plan and a short report.
What is a product discovery sprint?
It is the first step of a new product, before design and development. The work is to learn who has the problem, how they handle it today, what a first version would need to do, and whether it can be built in a sensible way. Everything in the sprint serves one question: is this worth building, and if so, what exactly?
The idea comes from the discovery phase used by product and service teams. The UK government's guide to how the discovery phase works puts it plainly: you should not start building your service in discovery. A sprint applies the same rule to a smaller question and a smaller budget, so it suits a founder or a business owner as much as a large team.
The length depends on the question. A narrow idea with easy access to users needs less time than one that depends on a payment provider, an old system or a regulator. Agree the length and the price at the start, and agree what would count as finished.
What should a discovery sprint decide?
A sprint exists to settle a business question. By the end, you should be able to answer five things in writing:
- Who the first user is: one group you can name and reach. "Small businesses" is too broad to recruit from.
- What problem you are solving for them, in their words, and what it costs them today.
- What the first version does and what it leaves out. The list of exclusions keeps the first build small enough to afford.
- How it would be built: the technical approach, the systems it has to talk to and the riskiest part of the build.
- What happens next: build, change direction or stop, with the reason.
The Nielsen Norman Group describes the same end point in its article on discovery in Agile: by the end, a team should be able to state the problem or opportunity and propose possible solutions with confidence. If a sprint cannot reach a decision, the uncertainty does not go away. It turns up later as changes during the build, when each change costs more.
What does the business need to prepare?
A sprint goes faster and says more when the business arrives ready. You do not need a finished idea. You do need these:
- A person who decides. The founder or product owner should be available at each checkpoint, because a sprint that waits a long time for a yes or no loses its focus.
- Access to the people who have the problem: customers, staff or prospects who will give twenty minutes to talk. If you cannot reach any, the sprint will find that out first.
- What you already know: past enquiries, complaints, sales notes, earlier attempts and the spreadsheets people use as workarounds. Old evidence is cheaper than new research.
- The constraints: the budget range for the build, any date you must hit, systems the product must connect to, and rules that apply, such as data protection for personal data.
- A view of what would make you stop. Write down, before the work starts, what result would mean the idea is not worth building, so nobody reads every result as encouraging later.
If you are unsure whether your idea is ready for a sprint, our guide on how to validate a product idea before building it covers the cheaper tests you can run on your own first.
What should you receive at the end?
This is where sprints differ most, so ask for the list before you agree to one. Each item below should be a document or file you keep after the sprint ends.
Handover | What it holds | Who owns it | What you use it for next |
|---|---|---|---|
Problem statement | The problem, who has it and what it costs them today | You | Briefing any team, investor or partner |
Research notes | What people said and did, in their words, with the questions asked | You | Product copy, onboarding and later decisions |
Scope of the first version | What is in, what is out, and why | You | Getting a build quote based on a clear scope |
Clickable prototype | Screens that behave like the product for the main task | You | Testing with users before any code, and showing investors |
Technical approach and build plan | How it would be built, the systems it connects to and the main risks | You | Comparing quotes and planning the build |
Report with a decision | What was learned and a plain recommendation: build, change or stop | You | Deciding whether to spend on the build |
Two things make this list useful. The first is ownership: every item should be yours to keep and to share with any team, including one you choose instead of the team that ran the sprint. The second is that each item leads to a next step. A scope you can hand to builders lets them quote on the same thing, which our note on writing a brief vendors can quote on explains in more detail.
Be wary of a report built on compliments. Before you spend on a build, it should tell you what people actually did when they tried the prototype, which questions are still open, and what new evidence would change the recommendation.
What if the answer is do not build?
Then the sprint has saved you the cost of the build. The UK government guide makes the same point about discovery: it is not a failure to stop if the research shows that is the best thing to do.
A "do not build" result usually takes one of three forms:
- The problem is real, but for a different group. Change the first user and test again.
- The problem is real, but people will not change how they work to get your solution. Look for a smaller change that fits their habits.
- The problem is not painful enough. Nobody is spending time or money working around it today, which makes this the clearest stop.
In each case the report should still be worth keeping. It records what you learned, so the next idea starts from evidence instead of from zero.
How is it different from a prototype or an MVP?
The three are often confused because a sprint can include a prototype and lead to an MVP.
- A discovery sprint answers whether to build and what. Its output is a decision and a scope.
- A prototype is clickable screens with no real software behind them. It tests whether people understand the product and can complete the main task. It is usually one part of a sprint.
- An MVP is working software that real users can use for one core job. It tests whether people use the product and come back, which a prototype cannot show.
If the problem and the first user are already clear and the only open question is whether people will use it, skip the sprint and build a narrow first version. Our guide on how much of an MVP to build first covers that step.
How we run it
Our Product Discovery Sprint follows the steps above and ends with every item in the handover list, a clear recommendation and a fixed price agreed before we start. You keep the prototype, the scope and the report, and you can take them to any team. When the answer is to build, the plan leads into Product Development or MVP development.
Have an idea you want to test before you pay for the build? Start with the Product Discovery Sprint.
Questions
Who from our side needs to take part?
One person who can make the final call on scope and budget, usually the founder or the owner of the product. It also helps to bring in someone who does the work the product will change, such as the person who handles orders or bookings today. They know the workarounds that never make it into a brief.
Do we own the research and designs if we build with another team?
You should. Before the sprint starts, get it in writing that the research notes, the prototype, the scope and the report are yours to keep and share. In our Product Discovery Sprint you get all of them and can take them to any team.
What happens if the sprint says not to build?
You stop, or you change the question and test again. That result is the sprint doing its job, because you learn it before the cost of a build. The report should still say what you learned and what would have to be true for the idea to work.
Is a discovery sprint worth it for a small internal tool?
Often not. If the people who will use the tool sit with you, the workflow is clear and the build is small and easy to change, talking to them and building a first version may be quicker. A sprint earns its place when the users, the problem or the technical route are still uncertain.
Filed under Product Development. The service behind it: Product Development.

