September 17, 2026
What Is Product Discovery? Definition, Risks and a Weekly Method
Product discovery is the work of deciding what to build before you build it. Learn the four risks, the methods that reduce them and a weekly discovery cadence.
TL;DR
Product discovery is the work that happens before delivery: finding a customer problem worth solving and collecting evidence that your solution is valuable, usable, feasible and viable. Interviews, usability tests, prototypes and a live feedback board are the core methods. Run it weekly, keep it separate from delivery sprints, and publish what you learn.
Author: Inputly
What is product discovery?
Product discovery is the work that happens before and alongside delivery: finding a customer problem worth solving, then collecting evidence that your intended solution is valuable, usable, feasible and viable. It starts from a problem, not from a feature list, and it ends when the evidence is strong enough to justify the build. Marty Cagan's four risks are still the cleanest checklist: will people choose it, can they work it out, can we build it, does it work for the business. Delivery is the other half, building the thing reliably. Teams rarely fail at delivery, they fail at building the wrong thing very well.
TL;DR: Discovery is problem first, evidence second, build third. Cover value, usability, feasibility and viability before you commit an engineer-month. Run a weekly cadence instead of one big research phase.
Why does product discovery matter?
- It stops expensive rewrites. A prototype test costs an afternoon, a shipped mistake costs a quarter.
- It gives prioritisation real inputs. Requests on a feedback board become opportunities with names and stories attached.
- It tells you what to say no to. Evidence makes saying no to feature requests a factual answer instead of a political one.
What are the four risks of product discovery?
| Risk | The question | Cheapest way to reduce it | | --- | --- | --- | | Value | Will anyone choose this? | Customer interviews, prototype tests, pre orders | | Usability | Can they work it out alone? | Usability testing on a clickable prototype | | Feasibility | Can we build and run it? | Technical spike, data or API check | | Viability | Does it work for the business? | Pricing, support load, legal review |
A bet is ready when all four have an answer you would defend in front of the company.
Which methods does product discovery use?
- Customer interviews. Ask about past behaviour, not about your idea. Good questions in user interview questions.
- Prototype and usability tests. Watch five people try to complete one task.
- A live feedback board. Continuous intake beats a research sprint, because it never stops.
- Surveys for signal, not decisions. Metrics such as CSAT flag where to dig, interviews explain why.
- Beta feedback or a customer advisory board, for narrow groups and the buyers who decide renewals.
How is discovery different from delivery?
| | Discovery | Delivery | | --- | --- | --- | | Goal | Reduce uncertainty | Ship quality work | | Output | Evidence and decisions | Releases and results | | Measure | What you learned | What you shipped | | Cadence | Weekly, continuous | Sprint or release cycle |
Keep the two separate on the calendar. Mixing them is how discovery becomes a kickoff call with a slide of assumptions.
How do you run a weekly discovery cadence?
- Open the week on the feedback board. Read the newest answers and merge duplicates.
- Book one 30 minute customer conversation. One is enough when it happens weekly.
- Test one prototype with five people, and write the findings the same day.
- Score opportunities before writing solutions: how many customers, how often, how painful.
- Publish what you learned and what changed, then answer replies on your public roadmap.
Frequently Asked Questions
- Does discovery delay shipping? Done weekly it speeds shipping up, because it removes the rework that unvalidated features create.
- Do you need a research team? No. A product manager, a designer and one engineer running regular conversations cover most early stage discovery.
- How do you know when to stop discovering? When the four risks have evidenced answers and the remaining questions are only answerable by shipping.
- What kills discovery fastest? Turning it into a slide deck. If nobody outside the product team hears a real customer voice, the evidence gets ignored.
- How does discovery connect to the roadmap? Discovery feeds it. Decisions backed by evidence earn a place, and the public version explains the why behind each theme.
Start with one interview this week and let the answers shape the next decision. Create a free space on Inputly to keep every answer, request and verdict in one place.