You do not need another person to say your idea sounds promising. You need to understand what happens when the problem appears and how they deal with it. A useful conversation leaves you with a clearer decision, even if that means changing your intended customer or dropping a feature.
This field guide offers a short interview structure and an evidence sheet. An interview does not establish demand on its own. It helps you design the next test of actual behavior.
Write the decision before recruiting interviewees
Imagine you are considering a tool for preparing sales proposals. This is a hypothetical example, not a Ramen Founders customer success story. Asking whether people like the idea provides little direction. Investigating how an independent consultant prepares a proposal under time pressure gives you a situation to explore.
Write down the person and context, the episode you want to reconstruct, and the decision you will make afterwards. You might be deciding whether an initial prototype should help produce one draft rather than manage an entire sales process. Leave room for an unfavourable answer.
Speak with people who have experienced that situation recently. Friends and contacts can introduce you to them, but friendliness is not a substitute for relevant experience. A small sample can uncover questions and patterns; it does not automatically represent a market.
Reconstruct an episode before pitching your solution
Eric Migicovsky's Startup School guidance focuses on specific situations and remedies people have already tried. Our practical suggestion is to turn the conversation into a sequence you can examine afterwards.
- Establish a recent episode: what the person needed to accomplish and what triggered the task.
- Walk through the steps, including waiting, repeated work and decisions.
- Explore the alternative they used and why it made sense at the time.
- Investigate the consequence of the difficulty: delay, effort, expense or abandoned work.
- Ask what additional context would help you understand the process.
Allow an answer to finish before explaining what you think it means. If someone says a task takes forever, ask for an episode that makes that description concrete. If they cannot recall one, preserve the uncertainty instead of filling the gap with the story you hoped to hear.
Keep observation separate from interpretation
After each interview, complete a five-field evidence sheet. You need a useful record, not an elaborate transcript.
- Context: the task and the circumstances around it.
- Observation: what was described or demonstrated, including any limitations.
- Current alternative: how the person manages today.
- Provisional interpretation: what you think it means and another plausible explanation.
- Next check: an action that could distinguish those explanations.
In the proposal example, seeing someone reuse an old document is an observation. Concluding that they need a new platform is an interpretation. A clearer template might be enough. Or the real delay might be obtaining information from their client.
Do not put another customer's information into an AI tool without permission. Remove identifiers when organising notes and retain only necessary context. You do not need to record an interview to learn from it; agree in advance if recording would help.
Agree on a proportionate next step
A friendly interview can justify further investigation. It does not establish willingness to buy. The next step could be reviewing an anonymized example, observing a task or agreeing to a bounded trial. Explain the actual offer and choose a commitment proportionate to what you have learned.
If someone declines, respectfully explore what makes the next step unsuitable. It could be low trust, another priority or a problem that matters less than you assumed. A refusal is information, not automatically an objection to overcome.
Your exercise for this week
Write one unresolved decision. Arrange three conversations as an initial learning round, not a scientific validation threshold. Complete the evidence sheet after each. Compare recurring behavior, contradictions and remaining gaps.
Then record a decision: narrow the segment, try a manual delivery or investigate another problem. Include the evidence that would make you reconsider.
Ramen Founders is a place to turn practical tools into useful work. Share the context of your project and one decision that changed after listening to someone. Protect the person's identity when describing the conversation. You can also explore our free founder tools.