Blog / Digital products /
A product validation interview guide for independent builders
You tell a few friends about your product idea. They love it. Several say they’d “definitely try it”. You hang up feeling brilliant.
Then you realise nobody told you how often the problem happens, what it costs them, or whether they’d change anything they do today. Kind words are not evidence.
A validation interview is a structured chat about someone’s current situation, workflow and decisions. It isn’t a pitch, and it isn’t a hunt for compliments. The aim is to test your assumptions and learn what should shape the next product decision.
Decide what you need to learn
Write down the main assumption before you schedule anything. Maybe a particular group hits a problem, uses a workaround, or would consider a new approach. Don’t try to validate audience, pricing, features and distribution in one conversation. I’d rather test one thing properly than five things badly, since a muddy answer helps nobody.
Talk to people who have lived the situation, not only people who know you. Define who fits based on the workflow, and don’t recruit only enthusiastic early adopters if the product is meant for a different group.
Plan a small set of open questions. You do not need a rigid script, but you should know which topics matter and what evidence would change your mind. Keep the interview focused and allow the participant to describe their own language and priorities.
Ask about past behaviour
Recent examples tell you more than hypothetical promises. Ask the person to walk you through the last time the problem came up: what started it, what they did, which tools they used and where it got hard.
Use neutral prompts such as “What happened next?” or “How did you decide?” Avoid leading questions like “Wouldn’t this feature save you time?” The wording can invite agreement without revealing whether the person would choose the product in real conditions.
Ask what they tried, what they stopped using and what the current workaround costs in effort or risk. Do not assume that a complaint means willingness to pay. People can dislike a process and still have higher priorities or acceptable alternatives.
Avoid pitching too early
If you show a prototype, first understand the current behaviour. Then explain what the prototype does without overselling it and ask the participant to complete a realistic task. Observe where they hesitate, what they expect and what they think will happen.
Don’t defend the design or correct every misunderstanding on the spot. Confusion is useful evidence, so ask what they expected and why. If the interface needs a long explanation from you, the product probably needs clearer onboarding or a different approach.
When asking about commercial commitment, use an honest and appropriate test. A positive reaction to a mock-up is not a purchase. Do not imply the product exists, is available or has capabilities that are only planned.
Record evidence responsibly
Take notes with permission and say how you’ll use them. Don’t collect personal or confidential details you don’t need. If you record, follow the consent and data-handling rules and store the file somewhere approved.
Keep observations apart from interpretation. “They reopened a spreadsheet” is an observation; “they need an automated dashboard” is a hypothesis. I’d rather have a messy page of honest observations than a tidy page of conclusions I invented afterwards.
Use consistent notes across interviews: context, current process, trigger, workaround, consequence, alternatives and unresolved questions. Do not turn an individual story into a broad market statistic. A small interview set can reveal themes and questions, not establish prevalence.
Look for patterns and contradictions
After several conversations, compare the problem and workflow. Do participants describe similar triggers? Are their workarounds alike? Do some groups consider the issue minor or solve it differently? Note both repeated patterns and meaningful differences.
Do not count every mention as equal evidence. A passing complaint differs from a problem that caused a participant to spend money, change tools or create a workaround. Still, strong past behaviour does not guarantee future purchase. Use it to decide what to test next.
Actively look for disconfirming examples. If people who fit the intended audience do not encounter the problem, ask whether your audience definition is wrong or your hypothesis is too broad. Avoid dismissing inconvenient feedback as a misunderstanding unless there is evidence.
Decide the next test
Summarise what you learned, what remains uncertain and what decision follows. You might narrow the audience, change the workflow, test a prototype, investigate pricing or stop the idea. Every option can be a useful outcome if it reflects the evidence.
Define a next test with a clear success or stop condition. If the issue is about repeated use, observe a realistic workflow. If it concerns payment, test a real commitment transparently. If it concerns discoverability, test a distribution route rather than adding another feature.
Keep a record of participants’ selection criteria, questions and limitations without overstating conclusions. This makes future product decisions more grounded and helps avoid repeating the same interview under a different name.
Treat the interview as a decision tool
Close by asking what they’d do next if the problem happened again, and what would make them change their current approach. Don’t ask them to design the product for you. Their behaviour and constraints are a better starting point than a wish list.
Afterwards, write down one assumption that got stronger, one that weakened and one that’s still open. That keeps the interview tied to a real product decision.
Related reading
- Planning a product launch around distribution, not only features
- A small product analytics plan that respects user privacy
- How to reduce support load with better product onboarding
A good interview replaces imagined behaviour with evidence about real decisions. For your next conversation, what past action would show whether the problem matters more than a polite “I like the idea”?