How to find the best features to build using the right customer interview questions
Tempo Team
Key Takeaways
The best customer interview questions in product management start from a written hypothesis about the problem.
Apply the 5 Whys inside the interview: Never stop at the first answer. The reasoning behind a workflow is usually more useful than the workflow itself.
Write the hypothesis and the disconfirming answer down before the interview starts. If the conversation cannot tell you which one won, the questions were not specific enough.
A customer interview in product management is about validating or invalidating an assumption the team has – listening closely enough to understand the customer's workflow and figure out what to build next.
Call it human nature: A PM walks into a customer interview with questions designed to confirm the thing they already want to build. They hear what they expect to hear, and leave with the same goals they entered with.
We're here to avoid that. Good customer interview questions work backward from a hypothesis about a problem, and here's how to approach the process effectively.
How many customer interviews are enough?
There's no fixed number, but most product teams get solid signal after talking with 5–7 users in each main segment (admins, day-to-day users, decision-makers). Interview until the same themes repeat – that's saturation. Include a mix of current, prospective, and churned customers for a balanced view.
For high-impact features or brand-new products, plan for more interviews and revisit them as your design evolves.
Ready to start building your own product roadmap? Check out our agile roadmap guide to see how insights from interviews connect to your planning process.
Questions to ask in customer interviews
Amy He, Product Manager at Google, recommends getting something tangible in front of users when you can:
"The best thing is to even give mocks to users to actually see how they'd use something – and really see what's confusing them and what's bringing them delight."
Always start with a hypothesis. Before any user research session, product managers need a working theory about what problem they're investigating. The goal is to validate or invalidate that assumption – not to fish for feature ideas.
Go in with an open mind. Your initial hypothesis might be wrong. If the customer's experience contradicts your assumption, that's valuable information.
Bring mockups when you have them. Once you've identified a problem to explore, show the customer a high-fidelity mockup. Ask them to accomplish a specific task while thinking aloud. Don't demo the product – just observe. Where they slow down, hesitate, or make errors tells you more than their stated opinions.
If you don't have mockups yet, use open-ended questions. The questions below give you a framework for uncovering problems worth solving – problems shared by enough of your target customers, problems that lead to sticky features, and problems whose solutions move your business metrics.
One practical note from Clement Kao at PMHQ: Most user interviews shouldn't go past 60 minutes. Interviewees run out of energy and focus by then.
Starter questions that uncover customer pain points
Start with 5 Whys. If your customer presents a list of feature requests or complaints, don't stop at the first answer. Ask why, then ask why again. Each level gets you closer to the root cause.
Take Henry Ford's famous (if apocryphal) example: "If I had asked people what they wanted, they would have said faster horses."
"Why do you need a new horse?" – It's old.
"Why do you need to replace it?" – It's too slow.
"Why does it need to go fast?" – I need to get from A to B efficiently.
Three whys in and the real problem emerges: Faster point-to-point travel. The horse was never the answer. Don't ask customers what they want. Ask about their process, their goals, and where they run into obstacles.
Rate the pain from 1 to 10. When a customer brings a list of gripes, ask them to rate each issue numerically. Which problems are actually painful? How often do they hit this wall? A number cuts through vague frustration and shows you where their priorities actually sit.
Questions about your users' current product experience
What led you to buy our product in the first place? This validates your product-market fit, your unique value proposition, and your assumptions about the target market. It also reveals how customers find you and what problem they were trying to solve when they did.
Why would you – or wouldn't you – recommend us to others? This is a win/loss conversation. It surfaces your differentiators, what's working, and what needs improvement. Keep asking why when something isn't landing.
What do you like most about our product? Find out what's working and why. There may be patterns in the answers you can replicate across other features.
What do you like least? (and then ask why) Listen for roadblocks. They point directly to where the product creates friction. "It takes too long to..." suggests automation would help. "I wish I could..." sounds like a feature request, but dig into the why before assuming. "It's so complicated that no one on my team will use it..." signals adoption problems – is there a simpler approach? Is more onboarding needed?
How are you using the product right now? Identify the most common and important tasks. Weed out ideas that don't address real use cases. Not every customer uses your product as intended – listen for out-of-scope use cases that might signal an adjacent opportunity.
How would you feel if X didn't exist? What do you think of doing Y instead? Gauge reactions to potential solution directions. Pushback is useful data.
Can you show me how you do that today? Watching someone walk through their current process shows you unnecessary steps, workarounds, and friction points they may not even recognize as problems. The status quo is often the biggest obstacle to adoption. A task that takes 10 steps today might take one with the right feature.
What do you like about how you do it now? User preferences can translate into design decisions. Something about the current workflow is working for them – don't throw it out without understanding what.
What do you wish you could do that you can't? This surfaces unmet needs and broken workflow steps. These answers often point to where your product can expand its value.
Questions that reveal your customers' goals
How do you measure success? Are your customers trying to save time, make money, reduce errors, or something else entirely? Understanding their success metrics helps you prioritize which features will actually move the needle.
How much would you pay for something that solved this problem? This nudges toward validation. Tells you how much they value the solution.
If you solved this problem, how much money would you save or make? Same purpose – quantifies the stakes and helps you understand the ROI your customers expect.
How would your job change if you had this? Draws out their underlying goals. If you've discussed several potential solutions during the interview, this question helps identify which one matters most.
The goal of customer interview questions
You're looking for the pain points that the largest group of your key users share. The most strategic product decision you can make is to build solutions tied directly to those problems and goals.
As you interview customers, they'll often confirm ideas you already had – or surface ones you didn't. Log those insights into a centralized location immediately for idea development, where you can track user problems alongside potential solutions before anything moves to the roadmap.
Tempo Strategic Roadmaps helps you capture, organize, and prioritize the ideas that come out of customer interviews – so insights don't get lost between the conversation and the next planning cycle. Try our product roadmap templates to get started.
Our guide to product ideas
Product managers draw on many sources of information when deciding what to build. Here's a guide to building actionable strategies for capturing and prioritizing the ideas that generate the most value:
Sign up for a demo
Request Demo
