How to Run Customer Discovery Interviews That Actually Tell You Something

Most customer discovery interviews end with the founder feeling encouraged and having learned almost nothing. The customer was polite. The founder asked questions that invited confirmation. Both people said the right things and left without anything real having changed.

This is the discovery interview failure mode. It is extremely common, especially for first-time founders who are conducting these conversations while managing a full-time job and an idea they have been sitting on for months.

The good news is that running an interview that produces real signals is a learnable skill. It requires a different starting point than most founders walk in with, and a few specific habits that are easy to apply once you understand why they work.

The Mindset Shift That Changes Everything

Most founders approach discovery interviews as confirmation exercises. They have an idea, they believe in it, and they want to find out whether other people believe in it too. This is understandable — and it almost always produces bad data.

When you are seeking confirmation, you ask leading questions. You describe the solution in the process of asking about the problem. You interpret polite engagement as a positive signal. And in a market like the UAE, where social courtesy is deeply embedded, polite engagement is exactly what you get — whether or not the person is actually interested.

The shift is from validation to diagnosis. A diagnostic call is not trying to confirm that the problem exists. It is trying to understand exactly what shape it has: how often it occurs, in what context, what it costs the person who has it, and what they currently do about it. Those are different questions, and they produce different answers.

Four Things to Do Differently

1. Ask about the last time, not about hypotheticals. “Would you use something like this?” is one of the least reliable questions in customer discovery. People are poor predictors of their own future behavior, and in this region especially, a polite yes rarely converts into money. Ask instead: “Tell me about the last time you ran into this problem.” Real memories are specific. If the customer can describe a particular moment, with details, the problem is real. If the answer is vague, the problem isn’t sharp enough in their life to matter.

   
   

2. Open with the problem, not the product. If you introduce your idea in the first five minutes, the conversation shifts from an inquiry to a pitch review. The customer is now evaluating your solution rather than describing their experience. Keep the product out of the conversation entirely until you have a clear picture of the problem. What you learn in that first half will tell you whether your solution is even addressing the right thing.

   
   

3. Listen for workarounds. If someone has a real problem, they have usually found an imperfect way to manage it. Ask what they currently do when the frustration occurs. The workaround tells you whether the problem generates actual behavior, and what the person is willing to invest to address it. Elaborate workarounds mean real problems. No workaround at all is a signal worth examining.

   
   

4. End with a small ask that costs something real. Not “Would you buy this?” but something more immediate and specific: “Would you introduce me to one other person who experiences this?” Or: “Can I follow up with you when I have an early version to show?” A yes that requires nothing is not a signal. A yes that costs the person time, a connection, or a small commitment is.

“A yes that requires nothing is not a signal. A yes that costs the person something — time, a connection, a commitment — is.”

What to Do With What You Learn

The point of a discovery interview is not to walk out feeling validated. It is to walk out with a clearer picture of the problem and the person who has it.

After each interview, write down three things: the exact language the person used to describe the problem, their current workaround, and the moment in the conversation when the energy shifted — when they leaned in, spoke faster, or said something they seemed surprised to say. Those moments are usually where the most useful information lives.

The pattern you are looking for across multiple conversations is not agreement. It is specificity. When multiple people describe the same problem in similar language without being prompted, the problem is real. When the memories are detailed and the workarounds are consistent, you have something to build toward.

How Many Interviews You Need

Five to eight honest conversations, conducted this way, will tell you more about your customer than months of desk research. You are not looking for a majority vote. You are looking for a pattern — the consistent shape that a real problem has when multiple people describe it independently.

Six people who describe the same frustration in similar terms, with specific memories, and at least some of whom take you up on a small ask — that is a meaningful signal. Six people who were kind and said the idea sounds interesting are not.

The FRWRDx IDEA Program’s first two milestones — Problem and Customer — are structured around exactly this process. If you want a guided framework for conducting discovery interviews that produce real signals, and a mentor network to help you interpret what you find, rolling applications are open. 14 weeks, AED 3,000. No equity.

Apply