Home/Blog/Product Discovery/How to Interview Users Without Leading T…
Product Discovery

How to Interview Users Without Leading Them: A Guide

How to interview users without leading them: ask about past behaviour, never pitch the solution, treat enthusiasm as weak evidence and trust small commitments.

Cover of the FIKR Space guide: How to Interview Users Without Leading Them: A Guide

To interview users without leading them, ask about what they do and what they have done, never describe your solution and ask whether they like it, and treat a small commitment as stronger evidence than a large compliment. Past behaviour is evidence. Predicted behaviour is imagination.

Conversations are one of the cheapest forms of discovery, and one of the easiest to waste. This post is about making them yield something you could actually act on.

How to interview users without leading them: start with the usual mistake

The most common way founders and product teams waste a user conversation is by describing the solution and then asking whether people like it. It feels natural. You are excited, you want a reaction, and the person across the table is polite.

That politeness is exactly what you get back. People are generous in conversation. They say it sounds useful, they say they would try it, and they mean it at the time. But their kindness in a conversation tells you very little about what they will do with their own time and money later.

So the fix is to change what you ask about. Stop asking for opinions of your idea and start asking for accounts of their life.

Ask about behaviour and history

The guide gives a short set of questions that all point backwards, at things that really happened.

Ask thisWhat it reveals
What do you do about this today?Whether the problem is real enough to have a workaround
When did you last do it?Whether it is a recent, recurring event or a hypothetical
What happened?The actual sequence, in their words, not your summary
What did you try before?How much effort they have already spent on the problem
What would have to change?What stands between them and a different behaviour

Notice that not one of these questions mentions your idea. That is deliberate. The aim is to learn about the problem, which is the first thing that must be true for any idea to work.

Free guide, 34 pages: The Dual-Track Playbook. Send me the guide.

Leading questions and their better versions

Here is how the same curiosity can be asked in a way that steers, and in a way that does not.

Leading habitBetter question, from the guide
Describing the solution and asking whether people like itWhat do you do about this today?
Asking whether they would use itWhen did you last do it, and what happened?
Asking for a verdict on the ideaWhat did you try before?
Asking them to predict their own behaviourWhat would have to change?

The leading versions invite a yes. The better versions invite a story, and a story can disagree with you. If a conversation could not possibly have told you that you were wrong, it was not a test.

Be suspicious of enthusiasm

An excited reaction feels like success. It is not. Enthusiasm costs the speaker nothing, and the guide's advice is to treat it as weak evidence.

What counts as stronger evidence is commitment, even a small one. The more the person gives up, the more the signal means.

  • A compliment costs the speaker nothing, so it says very little.
  • A commitment, even a small one, costs the speaker something, so it says more.
  • People are much less generous with their time and money than with their words, which is why spending either is the better signal.

This is why the guide says a small commitment beats a large compliment. If you leave a conversation with warm words and nothing else, record it as weak evidence. It may still be worth following up, but do not let it carry a decision on its own.

Speak to people who match the target

A conversation is only as informative as the person in it. Talk to people who match the target for the idea, not whoever is available. Colleagues, friends and anyone who is already fond of you will reliably tell you what you hope to hear.

This is a close cousin of one of the guide's warnings about declaring something validated on weak evidence: five friendly users, an internal opinion, a survey where everyone said yes. If the people you spoke to were friendly rather than representative, say so in your notes. The related post what does validated mean in discovery covers how to set the bar for what counts.

Decide what you are testing before you talk

An interview is a test like any other, so it should follow the same discipline.

  1. Name the assumption. Most interviews test whether people have the problem at all, which is usually the riskiest assumption of the lot. See how to test your riskiest assumption.
  2. Agree what a no looks like. If everyone describes a recent, costly experience of the problem, that is a yes. If nobody can recall the last time it happened, that is a no, and you must be prepared to hear it.
  3. Set a limit. A number of conversations or a date, before you start. It always feels responsible to learn a little more, and open-ended interviewing is how discovery drifts.
  4. Ask whether you would stop. If a negative picture would not change what you do, skip the interviews and build, which is at least honest.

Keeping the conversation on the past

A few reminders, all drawn from the guide, keep a conversation pointed at evidence and away from imagination.

  • Prefer history to prediction. When someone says what they would do, bring the conversation back to what they did, and when.
  • Do not pitch. Describing the solution and asking whether people like it is the move that reliably produces politeness.
  • Look for the next step. Ask what would have to change for them to behave differently, and note whether they offer anything beyond words.
  • Be ready to hear a no. A conversation that could not have told you that you were wrong was not a test.

None of this requires special training. It requires wanting to be told you are wrong, which is the real discipline underneath.

FIKR OS, strategy meets execution: OKRs, roadmaps and kanbans in one system. See FIKR OS.

Reading what you heard

After the conversation, sort what you collected by how much weight it can bear. A simple table keeps the sorting honest.

What you heardHow to treat it
Warm words about the ideaWeak evidence. Note it, but do not let it carry a decision
An account of what they did and whenEvidence. Past behaviour is the real material
A description of what they would doImagination. Ask when they last did something similar
A small commitmentStrong evidence. It cost them something to offer
A friendly source who does not match the targetDiscount it, and say so in the notes

The point of the table is not precision. It is to stop a pleasant hour from being recorded as validation. If the only things in your notes are in the first and third rows, the honest verdict is that you have not yet learned much.

When a conversation is the wrong test

Interviews are not the only cheap test, and sometimes they are not the best one. The guide lists others: a landing page, a prototype, a manual version of the service, a concierge run with a handful of customers, or a look at data you already have. Some assumptions, such as whether people will actually pay, are better tested by asking for something small than by asking for an opinion. Choose the conversation when the question is about the problem and the person's history with it, and choose something else when the question is about what they will really do.

Whichever you pick, the test is the same as for any discovery step: can the result change your mind, and will you actually stop if it is negative?

What to do with what you heard

After the conversation, the notes should answer the same four questions as any other discovery step: what we asked, what we did, what we found, what we decided. Keep that in the shared record, where someone who was not in the room can find it. The next post in the series on how to keep a discovery record explains why that is worth the few minutes it takes.

Share what you learned with the people who will build, too. They should be present in the discovery that shapes the work, not for every interview, but enough that the reasoning is shared and not merely passed on in a summary.

Free guide, 34 pages: The Dual-Track Playbook. Send me the guide.

A checklist for the next conversation

  • I know which assumption this conversation is testing.
  • I will ask about past behaviour, not predicted behaviour.
  • I will not describe the solution and ask whether they like it.
  • I will treat enthusiasm as weak evidence and commitment as strong.
  • The people I am speaking to match the target.
  • I know what answer would count as a no.
  • I will write down what we asked, did, found and decided.
FIKR OS, strategy meets execution: OKRs, roadmaps and kanbans in one system. See FIKR OS.

Frequently asked questions

Is it ever fine to show my idea in an interview?

The guide's concern is describing the solution and asking whether people like it, because that produces politeness. If you do show something, do it after you have learned about their history, and judge the reaction by what they commit to, not by how warmly they speak.

What if everyone I speak to loves the idea?

Be suspicious. People are generous in conversation and much less generous with their time and money. Look for small commitments, and treat the warm words as weak evidence until something stronger appears.

Who should I talk to?

People who match the target for the idea. Whoever is simply available is the wrong group, because friendliness and convenience are not the same as being the customer you are trying to serve.

How many conversations are enough?

Decide before you start, and tie the number to the decision. A small reversible change needs far less evidence than a quarter of engineering. State the bar in advance so it cannot quietly move.

The full method, with the checklists, is in The Dual-Track Playbook, a free 34-page playbook.

LG

Luis Goncalves

// Founder & CEO at FIKR Space

Three-time founder. Built and exited Evolution4All before this. Now building FIKR Space — the operating infrastructure underneath every innovation ecosystem (startups, accelerators, governments, investors). Lisbon-based, works global.

Claim your founding member spot.

Every founding member is hand-reviewed. Tell us who you are and how you want to co-create — we get back inside 48 hours.

Role
Job Title
Country
Cookies

We use cookies.

Essential ones keep the site working. The rest only run if you say so. No dark patterns, no pre-ticked boxes.Privacy policy

Co-create with us. It's about building it together.Co-Create With Us