Prepare a conversation, not a sales presentation
A useful invention interview starts with a learning objective. “Understand how small bakeries store cooling racks during a busy shift” is specific enough to guide questions. “Get feedback on my invention” encourages a demonstration followed by polite reactions. Write the three uncertainties you need to resolve and leave the prototype out of sight until the relevant problem discussion is complete.
NSF I-Corps Hub: Preparing for customer discovery interviews is a primary program resource on preparation. NSF: I-Corps customer discovery identifies discovery interviews as a source of market evidence. The script below is an original example to adapt, not a validated survey instrument or a requirement to ask every question in order.
A 25-minute interview structure
- Context and permission. Explain that you are researching the task, that participation is voluntary, and how notes will be used. Ask before recording. Let the participant decline questions.
- Recent episode. “Tell me about the last busy shift when you moved the cooling racks.” Ask when it happened and who was involved.
- Sequence. “What happened next?” Follow the task through setup, use, interruptions and cleanup.
- Friction and workaround. “Where did it become difficult?” Then ask what they did, why they chose that response, and whether it worked.
- Decision history. “What did you last buy or change to improve this?” Identify who approved it and what alternatives were considered.
- Summary and next step. Repeat the main finding in your own words, invite correction, and ask whether a later observation or fit test would be appropriate.
Keep the order flexible. If the participant mentions an important incident, follow it. A rigid questionnaire can interrupt the most useful evidence. At the same time, avoid turning the entire interview into a discussion of their unrelated business concerns.
Replace leading questions
| Avoid | Try instead | Why |
|---|---|---|
| Would a folding rack save space? | Show me where the racks go after use. | Reveals actual storage constraints. |
| How frustrating is the current system? | What happened the last time the area was full? | Allows a neutral or positive answer. |
| Would you pay $200? | How was the last equipment purchase approved? | Explores a real buying process. |
| Do you like this feature? | What would this change in your shift? | Connects a feature to consequences. |
A hypothetical interview turn
An inventor expects a bakery to need collapsible racks. The manager says, “Space is always a problem.” Instead of treating that statement as confirmation, the inventor asks about the previous Saturday. The manager explains that racks blocked a doorway because staff left them there while unloading, although a storage space was available nearby.
The issue may be workflow or placement rather than rack size. A follow-up question reveals that moving the empty rack requires two people because of a damaged wheel. The immediate alternative is a wheel repair, not a new product. This discovery is commercially valuable even though it weakens the original concept.
Later, another bakery describes a genuinely different pattern: off-site catering requires loading racks into a small van. Keep these findings separate. They point toward different segments and should not be averaged into a generic claim that “bakeries need folding racks.”
Handle opinions and contradictions
When someone says “everyone would buy that,” ask what they base the statement on. Record it as an opinion unless they can describe relevant evidence. When their words and behavior conflict, explore respectfully: “You mentioned that storage is difficult, but you also said the current rack is your preferred option. What makes it worth keeping?”
Do not argue with objections or explain why the participant misunderstood your design. If the explanation is necessary for safety or accuracy, give it, but note that you changed the conditions of the conversation. A persuasive demonstration and an unaided reaction are different observations.
Capture notes that can be used later
Separate direct quotations, observed facts and your interpretations. “Rack stood in doorway for six minutes” is an observation. “Staff hate current racks” is an interpretation unless supported by what they said. Assign an anonymous participant ID, record the role and context, and avoid storing unnecessary personal details.
Immediately after the interview, write the strongest surprise, the strongest contradiction and one question for the next interview. Do not wait until several conversations blur together. If you used an incentive, record it so you do not mistake willingness to participate for interest in the product.
The interview is successful when it changes a decision
A useful session may eliminate a feature, reveal a different buyer, or show that your problem statement is wrong. Decide what the evidence supports before scheduling more interviews. If you already have enough information to build a simple storage mockup, another ten general conversations may be less useful than that focused test.
Qualitative interviews are limited by recall, selection and social pressure. They do not establish market size or technical performance. Their value is in revealing real circumstances that your design and business assumptions must survive.