How to run customer interviews that tell you the truth

Why most founder interviews mislead

Most early interviews produce false encouragement because they ask people to predict the future.

"Would you use an app that…?" invites a polite yes. People are bad at predicting their own behaviour and good at being kind to an enthusiastic founder. The fix is to ask only about the past: what happened the last time the problem came up, what they did, what it cost, and what they tried before. Past behaviour is evidence. Opinions about a product that doesn't exist are not.

There's a second trap. Founders tend to interview people they already know, who want them to succeed. Friends are useful for practising the script. They're not useful as evidence.

Step 1: Write down who you think the buyer is

Before you contact anyone, describe the buyer narrowly enough that you could find them.

A useful description has four parts:

1. Role or household type. "Owners of independent dental clinics with one to three dentists", not "healthcare businesses".
2. Region. Where they are, and whether language matters. Buyers in Quebec should be interviewed in French.
3. Trigger. The moment the problem bites, such as month-end billing, a new hire, or the first snowfall.
4. Who pays. The person who would approve spending money, who may not be the person who feels the pain.

Write this on one line at the top of your notes template. You'll revise it as you go, and that's normal.

Step 2: Find twenty strangers who fit

Recruiting is the slowest part of discovery, so plan for three to five requests per conversation you get.

Places that tend to work in Canada:

• Industry and trade associations, including provincial chapters, which often have member directories or community boards.
• Online groups and forums where your buyers ask each other questions.
• Suppliers and adjacent businesses that serve the same customers, such as an accountant who works with many small clinics.
• LinkedIn searches by role and region, for business buyers.
• Local business networks such as chambers of commerce or business improvement areas.

Your request should be short, specific and honest. A template:

Hi [Name], I'm researching how [type of business] handle [problem area], and I'm not selling anything. Would you be open to a 20-minute call in the next two weeks about how you handle it today? I'll share a short summary of what I learn from everyone I speak to.

The offer of a summary matters. People often say yes because they're curious how their peers handle the same problem.

Step 3: Run the conversation

Each interview has five parts and should take about twenty minutes.

Part 1: Context (2 minutes)

Ask about their role and situation. "Tell me a bit about your business and what your week looks like." This relaxes people and gives you a check on whether they fit the buyer description.

Part 2: The last time (8 minutes)

This is the core. "Tell me about the last time [problem] came up." Then follow the story:

• "What happened next?"
• "Who else was involved?"
• "How long did that take?"
• "What did it cost, in time or money?"

Keep asking for the specific instance. If they slip into generalities ("we usually…"), bring them back: "Could you walk me through the most recent one?"

Part 3: What they've tried (5 minutes)

"Have you tried to fix this? What did you try? Why did you stop?" Workarounds are gold. A spreadsheet someone built, a part-time person they hired, a tool they pay for and dislike: each one proves the problem is real enough to act on.

Part 4: Priority (3 minutes)

"Where does this sit among the things you'd like to fix this year?" A problem that ranks fifth is not one people will pay to solve soon.

Part 5: Close (2 minutes)

"Is there anyone else I should talk to?" Then, only if they ask, a one-sentence description of what you're thinking about. Note their reaction.

Step 4: Log every interview the same way

A fixed notes template stops you remembering only the conversations that agreed with you.

Fill it in within an hour of each call, while the conversation is fresh.

Step 5: Read the patterns

After about twenty interviews, count rather than recall.

A worked example: a founder interviews 20 owners of small landscaping companies about scheduling. The tally reads:

• 18 fit the buyer description.
• 11 described scheduling trouble before it was named.
• 9 have a workaround: 6 use a whiteboard and texts, 3 pay for software they dislike.
• 7 rank it in their top three problems for the year.
• 4 asked when they could try something.

That's a strong signal: most fit, well over half raised it unprompted, half have workarounds, and several asked to try a solution. But the notes also show something the founder didn't expect: 8 of the 18 mentioned that their real cost was customers who don't pay on time. That's worth a second look before scoping anything.

A rough rule we use: if fewer than about a quarter of qualified people show strong signs of pain, such as raising it unprompted, paying for a workaround or ranking it highly, the problem probably isn't urgent for that group. Change the group or the problem before you build.

Common mistakes, and how to catch them

• Pitching. If you've described your idea in the first ten minutes, the rest of the interview is contaminated. Note it and discount that conversation.
• Leading questions. "Wouldn't it be great if…?" has only one polite answer. Ask "how do you handle…?" instead.
• Stopping early. Five interviews feel like a lot. They aren't. Patterns that matter tend to show between ten and twenty.
• Counting compliments. "That's a great idea" is not data. A workaround someone pays for is.
• Ignoring the surprise. The problem you weren't looking for is often more valuable than the one you were.

What to do with the results

The output of discovery is a one-page problem statement: who has the problem, how often, what it costs them, what they do now, and who pays to fix it. That page becomes the brief for scoping a first version, covered in the MVP cut-list guide.

If you'd like a second opinion on your notes, or want us to run a full customer discovery sprint with you, call, text or email. The first 30-minute consultation is free; pricing covers what comes after. More guides are listed on the guides page.