2-1
Customer interview questions to test a business idea
A short script built on what people did last time, the four questions that only get compliments, where to find ten strangers, and how to tally the answers.
- Written by
- Published
- Reading time
- 8 min read
- Sources
- 2 cited
- Time needed
- About a week for ten conversations
- What it costs
- Nothing, or the coffee
- Difficulty
- Tools required
- A notebook or spreadsheet for the tally Toolbox
Ask people in your group about the last time they had the problem (what happened, what they did about it and what it cost them), never whether they’d buy your idea. Talk to ten strangers, tally who raised the problem on their own and who already spends on it, and let that count decide whether you go on, change the group or stop.
In this guide
The customer interview questions that tell you something ask about the last time the problem happened, not about your idea: “Tell me about the last time you…”, “What did you do about it?”, “What did that cost you?” People are polite about ideas and honest about what they did last week. Below is a script of about ten questions in the order to ask them, the four questions to drop, where to find ten strangers in a week, and a tally sheet that turns ten conversations into a decision: go on, change the group, or stop.
The rule: ask about last time, not about your idea
A customer interview isn’t a pitch, and it isn’t a vote on your idea. Its job is to find out whether the problem is real, frequent and costly enough that people already spend money or hours on it. If they do, there may be room for a business. If they don’t, no amount of enthusiasm for your idea changes that.
Rob Fitzpatrick’s book The Mom Test is built on this rule: talk about their life rather than your idea, and ask about specific things they’ve done rather than what they think or might do, so that even someone who loves you can’t hand you a kind lie. Y Combinator teaches the same in its Startup School: the recap of Eric Migicovsky’s talk on talking to users names the three common mistakes as talking about your idea, asking about hypotheticals and talking too much.
In practice:
- ask about the past, which happened, not the future, which is a guess;
- ask about one specific time, not what they “usually” or “ever” do;
- listen for money and hours already spent, because that’s what the problem is worth to them.
Who to talk to and where to find them
Talk to strangers in one named group who’ve had the problem recently, say in the last three months. Leave out friends and family: they’re kind, and they’re rarely the group. They’re useful for one thing, though: introductions.
Where to find ten in a week:
- Introductions: ask everyone you know, “Who do you know who runs a small salon?” (or whatever your group is). One name each adds up fast, and an introduced stranger says yes more often than a cold one.
- Where the group gathers online: trade forums, subreddits, and Facebook, Slack or Discord groups for the trade or the situation. Read each group’s rules first; many ban promotion, so ask a moderator before you post a request.
- Where they gather in person: trade association meetings, local business groups, markets, meetups and conferences.
- Where they work: for local businesses, turning up at a quiet hour often beats an email. YC’s recap mentions a startup whose founders simply visited fire stations.
Keep the request short and honest, and make it about them:
“Hi [name], I’m looking into how [small salons handle no-shows]. Could I ask you about the last time it happened? Fifteen minutes by phone, any day this week. I’m not selling anything.”
A booking link saves the back-and-forth over times; free scheduling and form tools are in the Toolbox.
The script, in order
Plan on fifteen to twenty minutes. Swap the brackets for your group’s problem.
The opening
Thank them, say what you’re looking into in one line, and say you’ll ask about their own experience, not show them anything. Then start with the first question and let them talk.
The questions
- “Tell me about the last time you [dealt with the problem].” Let them tell it in full; everything else hangs on this story.
- “What happened next?” Keep them on that one time, step by step.
- “What was the hardest part of that?”
- “Why was that hard?” The answer is often a different problem from the one you came in with, and sometimes a better one.
- “When was the time before that?” This gives you how often it happens from facts, not a guess.
- “What did it cost you: money, hours, lost work?” Get the numbers.
- “What have you done to try to fix it?” “Nothing” is an answer, and an important one.
- “What do you use for it now, and what does that cost?” Count spreadsheets, staff, a friend who helps and doing without.
- “What don’t you like about what you’ve tried?” That’s the gap a new business gets in through.
- “Who else deals with this, or decides what gets bought for it?” The person with the problem isn’t always the one who pays.
- “Who else should I talk to about this?” Each conversation should lead to the next one.
Questions 1, 3, 4, 7 and 9 are adapted from the five questions in YC’s talk.
The close
Thank them and ask whether you can get back in touch when you have something to show. A yes is a small commitment, and a sign the problem matters to them; a shrug is a sign too. Note either one.
Questions that only get you compliments
Four questions turn up in most interview lists and collect polite answers instead of facts. Each has a better question to ask in its place:
| Instead of | Ask | Why |
|---|---|---|
| “Do you think it’s a good idea?” | “Tell me about the last time you dealt with [problem].” | An opinion on an idea is free to give, and kind by default. |
| “Would you buy it?” | “What do you use for it now, and what does that cost?” | A yes about the future costs nothing; money already spent is real. |
| “How much would you pay for it?” | “What did the last fix cost you?” | A guessed number isn’t a price; a price said out loud, or a pre-order, tests one. |
| “Would you use it if it did [X]?” | “What don’t you like about what you’ve tried?” | Wishes for features pile up; complaints about real tools show what matters to them. |
If you catch yourself describing the idea, stop and go back to their last time. And if someone says “I’d definitely buy that”, take it as a compliment, not a customer: thank them, and ask for something that costs them a little, such as an introduction to someone else with the problem.
Running the conversation
- Keep it to fifteen or twenty minutes, by phone, video call or in person.
- Take notes, or bring someone to take them; YC suggests a cofounder or a friend, or asking whether you can record. Always ask before you record: in some states everyone on the call has to agree.
- Write down their words, not your reading of them, and every number: dollars, hours, times a month.
- Let silences run. The useful part often comes after a pause.
- When they generalize (“I usually…”, “I’d probably…”), bring them back: “And the last time?”
- Keep your idea out of it until the end, and mention it then only if they ask.
- Fill in the tally sheet within the hour, while you can still read your notes.
Tallying ten conversations
One row per person, one column per thing you need to know. Copy it onto paper or into a spreadsheet before the first call:
| # | Had it in the last 3 months | Raised it before you asked | Spends money or hours on it now | What they use now | Agreed to a next step |
|---|---|---|---|---|---|
| 1 | |||||
| 2 | |||||
| … | |||||
| 10 | |||||
| Yes | /10 | /10 | /10 | /10 |
Then read the counts. The cut-offs below are this guide’s working numbers, not research findings; write your own at the top of the sheet before the first conversation, so the results can’t move them.
- Go on to the next test if most of the ten (six or more, say) had the problem recently and raised it themselves, and several (three or more) already spend money or real hours on it.
- Change the group if the problem is strong for a few people who have something in common: a size of business, a trade, a stage of life. Run ten more conversations with only that group.
- Stop, or change the problem, if most never had it, or have it and spend nothing on it. A problem nobody spends on is one few will pay to fix.
What to do with the answers
- If it’s a go, the next tests are the price said out loud, a one-page offer and a pre-sale. How to validate a business idea before you spend money puts them in order with a pass mark for each, and how to pre-sell a product before you make it covers taking money for something that isn’t made yet.
- Keep their words. The way people described the problem is the wording for your offer page and your first messages, and it beats anything you’d write from scratch.
- Keep the people who said yes to a next step. They’re the start of your list for how to get your first customers for a new business, which begins with ten named people.
- Grade yourself honestly. Answer the demand questions in the Idea Inspection with what you heard, not what you hoped to hear.
The Validation section has the other tests, for when the conversations say go.
Questions
How many customer interviews do I need?
Ten strangers from one group is a good first round: enough to see whether most of them have the problem, few enough to finish in a week. If the answers keep surprising you, do more; once you hear the same story again and again, you’ve heard it. Y Combinator’s Startup School makes the point that good user research can start with as few as one or two people.
Should I tell people about my idea during the interview?
Not until the end, and only if they ask. Once people know the idea, their answers bend toward being kind about it, and you lose the plain account of what they did last time. If they ask early, say you’ll explain at the end; whatever they say about the pitch is a compliment, not evidence.
How do I find strangers to interview?
Ask everyone you know for one introduction to someone in the group, then go where the group already gathers: trade forums and online groups (read their rules on promotion first), trade associations, local business events and markets. For local businesses, turning up at a quiet hour often works better than a cold email. Keep the request short, honest and about their experience, not your product.
Should I pay people for a customer interview?
Usually not at this stage. People who have the problem often like talking about it, and payment can bring in people who want the money rather than people with the problem. For busy professionals a small thank-you, such as a coffee or a gift card, is reasonable; offer it up front and make clear it doesn’t depend on what they say.
Can a survey replace customer interviews?
No. A survey can only ask the questions you already thought of, and it collects opinions, which cost nothing to give. Run the conversations first to learn what to ask, then use a short survey to count how common those answers are across more people.