How to scope a discovery phase that leads to a real decision
Discovery should end with clear scope, technical direction and a credible estimate — not a slide deck that restates the brief.
A discovery phase is a paid, time-boxed piece of work that turns an idea or a vague request into an evidence-based plan. Done well, it answers four questions before significant design or engineering spend begins: what problem is worth solving, for which users, what the smallest viable solution looks like, and what it will take to build.
Three angles at once
Good discovery looks at the problem from business, user and technical perspectives simultaneously. Business alignment defines the objective and constraints. User research replaces assumptions with evidence. Technical discovery identifies the integrations, data and architecture risks that would otherwise surface halfway through the build.
Typical outputs
- A product brief and problem statement
- Research findings and defined user roles
- Current and future-state user flows
- A prioritised scope with explicit exclusions
- Wireframes or a clickable prototype of the riskiest flow
- Architecture direction, risks and dependencies
- A roadmap, team plan and budget range
How long it should take
Duration follows uncertainty. A narrowly bounded question might need a week or two. A new SaaS product or a significant redesign more commonly needs two to six weeks. Products with complex integrations, multiple user roles, AI features or compliance requirements need longer.
What discovery should not become
Discovery is planning and risk reduction, not a disguised build. It shouldn't attempt final UI for every screen, a fully coded design system or production integrations. Equally, it shouldn't end with a precise build quote while major technical unknowns remain — the honest next step might be a short technical spike, more research, or a smaller first release.
A simple completion test
Discovery is finished when the team can say, with evidence: who the primary user is, which workflows are in the first release and which aren't, what was tested, which assumptions must hold, and what it will take to build. If those answers exist, the next investment is a deliberate decision rather than a guess.
The purpose of discovery isn't documents. It's enough shared evidence that the next decision is deliberate.
