Research serves decisions
Research that does not feed a decision is a hobby. Before you recruit anyone, write the decision: "Should we build X or Y first?" "Why do trial users not activate?" Then write what you would need to learn to decide.
Pick the right method
| Question | Method | Output |
|---|---|---|
| Who are our users and what do they need? | User interviews | Needs, contexts, language |
| Can people use this design? | Usability tests | Problems ranked by severity |
| What do people do in the product today? | Analytics and session review | Behavior patterns |
| How many people feel this way? | A short survey | Rough size of a pattern |
NN/g notes the difference clearly: interviews capture attitudes and needs, often before a design exists. Usability tests observe behavior with a design in hand. You usually need both.
A two-week plan
Frame
Write the decision, the questions, and who you need to hear from.
Recruit
Reach out to customers, leads, and your network. Offer a small thank-you.
Run sessions
Five to eight sessions of 30 to 45 minutes, recorded with permission.
Synthesize
Cluster notes into themes, with a quote for each.
Share
A one-page summary with decisions and next steps.
Ask better questions
- Ask about past behavior, not future intentions: "Tell me about the last time you..."
- Use open questions and follow up with "Why?" and "Can you show me?"
- Avoid leading questions like "Would you use a feature that...?"
- Let silence do some of the work.
Make it a habit
Book recurring slots, such as two sessions every other week. A steady drip of customer contact keeps the team honest and removes the need for large, rare studies.



