A discovery interview should change a product decision. Focus on a recent workflow, the people involved and the evidence you still need before building.
Recruit around an event
For a fictional proposal-management tool, recruit someone who recently chased an approval. A procurement lead and a daily user may describe different obstacles. Tell participants how notes will be used and obtain permission before recording. A convenient friend can help rehearse the script, but is not automatically representative of the buying audience.
Use a flexible discussion guide
- Context: which proposal did you handle most recently?
- Sequence: walk through the handoffs using a redacted example.
- Breakdown: where did work stop, repeat or move outside the system?
- Workaround: what happened when the normal process failed?
- Buying process: who could approve a change and who might reject it?
Show your concept after exploring the current workflow. Leading with a feature demonstration can turn discovery into a polite usability review. When someone requests automation, ask what decision or error it would address.
Separate observation from interpretation
| Evidence | Working explanation | Next check |
|---|---|---|
| Several proposal versions circulate | Approval ownership is unclear | Observe one complete approval |
| Sign-off happens by email | A new account creates friction | Test a guest approval flow |
| An assistant follows up daily | Chasing consumes operating capacity | Record the actual follow-up workload |
How many sessions are enough?
There is no universal interview count that validates a business. Compare different roles, company sizes and purchase constraints. A small homogeneous sample can make a weak assumption look consistent. Keep contradictions visible instead of averaging them away.
The output is a short evidence log: what happened, what you infer and what you will test. An enthusiastic reaction belongs in the notes, but it is not equivalent to a pilot commitment or evidence of repeated use.
When interviews cross operating regions
Interview a local operator and a central buyer separately before treating their answers as one requirement. A headquarters purchasing process may hide branch-level workarounds. Keep the differences visible in the evidence log rather than describing a single global user.
Sources and further reading
From this guide to your project
Strategy ↗
An idea becomes an opportunity when it solves a real problem for people willing to change how they work.
Tell us about your ideaExamples are educational and do not describe client results. Content prepared with AI assistance; see our editorial policy. Our method ↗