How to Write Customer Discovery Questions That Reveal Real Problems
Write customer discovery questions that uncover real behavior, not polite opinions. Five example questions, what to avoid and how to read the answers honestly.
On this page
Customer discovery is the work of finding out whether a problem is real, who has it and how they deal with it today, before you build the solution. The quality of what you learn depends almost entirely on the questions you ask.
Ask the wrong way and people will tell you what they think you want to hear. Ask the right way and they will tell you what actually happened. This guide shows how to write questions that do the second, with five example questions you can adapt for an interview or a survey.
Why leading questions get you comfortable lies
A leading question hints at the answer you want. “Wouldn’t it be great if you could book a cleaner in one tap?” There is only one polite answer, and it tells you nothing.
People aren’t lying on purpose. They are being kind, trying to be helpful or imagining an idealized version of themselves. Ask “Would you exercise more if it were easier?” and almost anyone will say yes, regardless of what they actually do.
Leading questions are easy to write by accident. Watch for these signals:
- The question mentions your solution.
- It contains a positive or negative judgment (“frustrating,” “great,” “annoying”).
- It can be answered with a simple yes, and yes is the obvious answer.
- It asks people to predict their future behavior.
Ask about what people did, not what they would do
The single most useful rule in customer discovery: past behavior is evidence; future intentions are guesses.
Instead of “Would you pay for a tool that tracks your receipts?”, ask “Tell me about the last time you had to find a receipt for an expense. What happened?”
The second question gets you a story. Stories contain details you can check: how long it took, what tools they used, whether they gave up, whether it cost them money. If someone can’t remember the last time the problem occurred, that is a finding too. The problem may not be as frequent or painful as you assumed.
Good behavior questions usually start with:
- “Tell me about the last time…”
- “Walk me through how you…”
- “What did you do when…”
- “How often in the past month did…”
Investigate the alternatives they already use
Nobody is waiting with a problem and no solution. They are doing something, even if it is a messy spreadsheet, asking a friend or ignoring it.
Existing alternatives tell you three things:
- How much the problem matters. If someone has built an elaborate workaround, it matters. If they shrug, it may not.
- What you are really competing with. Often it is not another company; it is a habit.
- What they would need to switch. Their complaints about the current method are your opportunity; their reasons for staying are your obstacles.
Useful questions here include “What have you tried to solve this?” and “What do you like about the way you do it now?” Don’t skip the second one. People stick with imperfect solutions for good reasons, and you need to know them. Those reasons often turn into customer objections later.
Questions to avoid, and what to ask instead
| Avoid | Why it fails | Ask instead |
|---|---|---|
| “Would you use an app that…?” | Asks for a prediction and pitches your idea | “How do you handle this today?” |
| “How much would you pay for…?” | People guess, and the guess is unreliable | “What do you currently spend on this, in money or time?” |
| “Don’t you hate it when…?” | Leads to agreement | “What is the hardest part of…?” |
| “Do you like this feature?” | Invites politeness | “When did you last need something like this? What did you do?” |
| “Would you recommend us?” (before launch) | There is nothing real to recommend yet | “Who else do you know who deals with this?” |
Also avoid double questions (“How do you plan and pay for meals?”). Ask one thing at a time, or you won’t know which part the answer refers to.
Five example customer discovery questions
Imagine you are exploring a tool to help freelance photographers manage client contracts and deposits. Here are five neutral questions, each with what it is designed to reveal.
-
“Walk me through what happened with your last new client, from first message to getting paid.” Reveals the real workflow, where the friction is and whether contracts even come up unprompted.
-
“When did a client last pay late, cancel or dispute something? What did you do?” Tests whether the problem has actually occurred recently and what it cost them.
-
“What do you use today to handle contracts and deposits, and how did you choose it?” Uncovers the existing alternatives and what mattered when they made that choice.
-
“What do you like about your current way of doing it?” Reveals the reasons people stay put, which will become your biggest switching obstacles.
-
“In the past year, have you spent money or time trying to improve this? What happened?” Separates people who feel the problem enough to act from people who merely agree it is annoying.
Notice that none of these mention your tool. In an interview, follow each with “Why?” or “Can you tell me more about that?” The best information usually comes in the follow-up. In a survey, pair open questions like these with a few multiple-choice ones about frequency and current tools, so you can compare answers across people.
How to interpret feedback without overstating it
Collecting answers is the easy part. Reading them honestly is harder, because you want the idea to work.
- Count stories, not compliments. Five people describing a recent, specific struggle is stronger than twenty saying “sounds useful.”
- Note who said what. If the strongest pain comes from one specific segment, that segment may be your real customer.
- Look for actions. Did anyone ask when they could try it, offer to pay or introduce you to someone? That is the closest an interview gets to a real signal.
- Keep small samples in proportion. Ten conversations can reveal patterns and surprises. They can’t tell you how common something is across a whole market.
- Write down what would disprove your idea before you start, then check honestly whether you heard it.
Simulated audiences can help you prepare. Running your concept past AI-simulated participants can suggest which segments and doubts to probe in interviews. But simulated reactions are not discovery; they don’t replace a single real conversation. The guide on AI-simulated audiences vs. real customer research explains where each fits.
Your next step
Write down the problem you believe exists, then draft five questions that ask about past behavior and current alternatives without mentioning your solution. Read each one aloud and remove anything that leads, judges or predicts. Then schedule five to ten short conversations with people who fit your target customer. If you’d like a head start, generate a neutral draft and edit it to fit your audience.
Written by the SellTo.us Team. Examples in this article are hypothetical. SellTo.us results come from AI-simulated participants, not real people, and don’t predict sales or guarantee success.
Related articles
How to Find Customer Objections Before Launching a Business
Every new offer meets hesitation. Here is how to find the reasons people might say no before launch day, check which ones are real and fix them in your offer and message.
How to Roast Your Startup Idea Before Spending Money
A practical way to pick apart your own business idea before you commit savings, time or a team to it, and to turn every weakness you find into a small experiment.
AI-Simulated Audiences vs. Real Customer Research: What's the Difference?
Simulated audiences are fast and cheap, but they are not people. Here is an honest look at what they can and can't tell you, and how to combine them with real research.