What a discovery phase is
You have an idea, some budget, and pressure to start building.
Someone suggests a discovery phase first, and now you are weighing weeks of research against the pull of shipping something.
This guide helps you make that call: what a discovery actually produces, when the investment makes sense, and when you can reasonably skip it.
A discovery is a preliminary phase in which you research the problem space and frame the problem to be solved.
You gather enough evidence and initial direction to decide what to do next.
The Nielsen Norman Group describes it as work that sets a project off in the right direction by focusing on the right problems Nielsen Norman Group.
That way, the team builds the right thing.
Importantly, discovery is not solution testing.
It typically does not involve testing a hypothesis or evaluating a potential solution.
It is open-ended investigation: you are learning about the problem, not validating your answer to it Nielsen Norman Group.
If that distinction feels abstract, our discovery phase glossary entry breaks the term down with a concrete example.
What a discovery should produce
A well-run discovery leaves you with four things, according to the Nielsen Norman Group Nielsen Norman Group:
- Understanding of users. Who the users are, how the problem affects them, and what they need, desire, and value from a solution.
- Understanding of the problem and opportunities. How and why the problem occurs, its effect on users and the organization, and its magnitude.
- Understanding of constraints. What current business processes, existing solutions, and compatible technologies allow, so you know which solutions are feasible to explore afterward.
- A shared vision. Agreement with stakeholders on business objectives and desired outcomes, including an idea of what to measure going forward.
The payoff, when done well, is that solutions proposed later are desirable to users, viable for the organization, and feasible with the available technology .