Customer Discovery Questions That Actually Work

Why Most Founders Ask the Wrong Questions

I've done this five times now - built a company, scaled it, sold it, repeat. Every single time, the founders I watch struggle the most are the ones who skip or rush customer discovery. They fall in love with their own idea, do a handful of conversations that confirm what they already believe, and then spend six months building something nobody actually wants to pay for.

The data backs this up. 42% of startups fail because they build products nobody wants - that's the single largest category of startup failure, ahead of running out of cash, getting outcompeted, and every other cause. That number isn't random. It's what happens when founders treat customer discovery as a formality instead of the most important research they'll ever do.

Customer discovery isn't complicated. But it requires a specific kind of discipline: you have to genuinely not care whether your idea is good or bad. You're not there to pitch. You're there to learn. The moment you start steering the conversation toward validation instead of truth, the whole thing breaks.

These questions below are what I'd ask if I were starting from zero today. Some are classics from the startup world. Some come from the way I approach outbound - where you learn just as much about a prospect's real problems from a 10-minute cold call as you do from a formal interview.

What Customer Discovery Actually Is (And Isn't)

Before we get into the questions, let's get the definition straight - because I see founders confuse customer discovery with a lot of other things.

Customer discovery is the process of validating your assumptions about a target customer's problems, behaviors, needs, and decision-making processes before you invest significantly in building a solution. It is not a focus group. It's not a survey. It's not asking your LinkedIn network if they like your idea. It's structured, one-on-one conversation designed to surface how people actually experience a problem - their workarounds, their costs, their frustrations - before you've committed to building anything.

The key distinction that Rob Fitzpatrick nails in The Mom Test is this: good customer discovery questions are ones that even a supportive, well-meaning person couldn't lie to you about. The discipline traces back to Steve Blank's customer development framework - the idea that you shouldn't be building a product, you should be building a customer. The product comes after you understand the customer deeply enough to know exactly what they'll pay for.

What it isn't: a pitch meeting in disguise. The moment you introduce your solution into the conversation, you've changed the entire nature of the exchange. The customer stops telling you how the world actually is and starts evaluating whether your solution sounds good. Those are completely different conversations - and only one of them generates useful information.

The Foundation: What You're Actually Trying to Learn

Before you open your mouth, get clear on what you need to know. There are three things customer discovery should answer:

• Does this problem exist at all, and is it frequent enough to matter? A problem someone experiences once a year isn't the same as one they fight with every week. The best startups look for problems that customers face on a regular basis, or that are painful enough to create urgency on their own.
• Is there budget attached to solving it? Empathy and willingness to pay are very different things. People will tell you they'd pay anything to solve a problem, then not open their wallet. You need evidence of actual spending - either on your category or on the messy workarounds they've built to cope without a real solution.
• Who actually owns this problem inside the organization? In B2B especially, the person who feels the pain and the person who signs the check are almost never the same person. If you build for the user without understanding the buyer, you'll have fans but no revenue.

If you walk out of an interview without clear answers to all three, you wasted everyone's time. That's what the questions below are designed to fix.

Setting Up the Interview the Right Way

Most people rush past the setup and wonder why their interviews produce garbage data. The setup determines everything about the quality of what you get back.

Frame it as research, not a sales call. When you reach out to request the conversation, be explicit: you're doing research before building anything, you're not selling, and you genuinely want their honest experience. This framing matters because it sets the person's expectations. They'll be more candid when they know they're not about to be pitched.

Talk to strangers, not friends. One of the most common customer discovery mistakes is talking to people who want to be nice to you. Friends, former colleagues, and warm contacts who respect you will tell you what you want to hear. Cold outreach for discovery interviews is actually an advantage here - strangers have no social incentive to flatter you. Their feedback is honest by default.

Target the right people from the start. This seems obvious but most founders are sloppy about it. If you're building a B2B product, you need to be talking to the actual decision-makers in the right company size and industry - not proxies or people adjacent to the problem. Twenty interviews with the wrong segment teaches you nothing useful. To build a list of the right contacts fast, a tool like ScraperCity's B2B lead database lets you filter by job title, seniority, industry, company size, and location so you're targeting the actual decision-makers, not their assistants or adjacent roles.

Set the agenda upfront and keep it short. Ask for 20-30 minutes maximum. Lead with: "I'm going to ask you about your experience with [problem area]. I won't be pitching anything. I just want to understand how you currently handle this and what frustrates you most about it." That's it. That's the whole intro.

Record everything (with permission). Don't rely on memory or even real-time notes if you can avoid it. You want to be fully present in the conversation, and note-taking splits your attention. Ask for permission to record at the start. Most people say yes if you've framed it as research.

Opening Questions (The First 5 Minutes)

Your job at the start is to get the person talking about their life and workflow - not about your idea. Never mention what you're building until the very end of the conversation, if at all. People are polite. If you tell them your idea early, they'll spend the rest of the interview trying to be encouraging instead of honest.

• "Walk me through how you currently handle [the problem area] today." This is your anchor question. It opens the door to their actual workflow, not an idealized version of it. Let them talk for as long as they want here.
• "What does a typical [day/week] look like for you in this area?" Context matters. You want to understand where the friction lives inside their real routine, not a hypothetical.
• "When did this first become a problem for you?" The origin story reveals a lot - how long they've been tolerating it, what triggered awareness, and whether this is an evolving pain or a static one.
• "Tell me about your role and how this area fits into your day-to-day responsibilities." Understanding their role helps you figure out whether this person is the user, the buyer, or both - and calibrate everything else accordingly.
• "What were you doing before to handle this, and what changed?" Evolution questions surface transition moments. If they switched tools or approaches, there's almost always a trigger event worth understanding.

Pain Depth Questions (The Core of the Interview)

This is where you separate a real problem from a mild annoyance. You're looking for pain that is both frequent and intense. Frequent means it happens regularly enough to matter. Intense means it costs them time, money, relationships, or reputation.

• "What's the most frustrating part of dealing with this?" Open-ended, no leading. Let them name it.
• "How much time does this eat up for you in a given week?" Time is the first proxy for pain severity. If they can't quantify it at all, the pain may not be that sharp.
• "What happens if you don't solve this - what does the downstream look like?" This question reveals consequences. Strong answers here signal a real buying opportunity. Weak answers signal a nice-to-have.
• "On a scale of 1-10, how much of a priority is fixing this for you right now?" Anything below a 7 is a polite lie. Below a 5 means it's not actually a problem worth building for.
• "Has this ever cost you a deal, a customer, or a hire?" Specific, high-stakes consequences are the clearest signal that the pain is real.
• "Tell me about the last time this came up for you. What happened specifically?" This is the Mom Test principle in action - ask about specific past events, not future hypotheticals. People are bad at predicting future behavior. They're much better at describing what they already did. Behavior is evidence. Opinions are cheap.
• "How often does this come up - is it a daily thing, weekly, monthly?" Frequency matters as much as intensity. Pain that hits daily is far more compelling to build for than pain that surfaces quarterly.
• "What would happen to your business if this problem disappeared tomorrow?" Flipping the question to the positive can unlock different answers. If someone can't articulate a meaningful upside, the problem isn't as critical as it seemed.

Current Solution Questions (Your Competitive Intel)

What people already use to solve the problem tells you more than almost anything else. If they've cobbled together five tools and a spreadsheet, there's a gap. If they tried a competitor and quit, you need to know why.

• "What are you using right now to handle this?"
• "How did you end up with that solution?"
• "What do you like about it? What drives you crazy about it?"
• "Have you tried anything else before this? What happened?"
• "If you could wave a magic wand and change one thing about how you deal with this today, what would it be?" This last one is gold. The answer is almost always your headline feature.
• "How long have you been using this approach? Have you ever seriously considered switching?" Switching inertia is a real competitive moat. If they've been on something for three years and never considered switching, you're not just building a better product - you need a much stronger wedge.
• "What would a perfect version of this look like for you?" Let them describe the ideal state in their own words. Their description is your product spec and your marketing copy, handed to you for free.

One thing I see founders miss constantly: the competitor question. People won't always volunteer what they're paying for alternatives. Ask directly: "What are you spending on this problem today, between tools, time, and any outside help?" That number is your pricing anchor.

The Urgency and Timing Questions (Often Skipped Entirely)

Most founders ask about pain. Few ask about timing. But urgency is what separates a customer who buys this quarter from one who says "send me something" and goes silent for six months.

• "What would need to be true for you to prioritize solving this in the next 30 days?" This question surfaces the trigger conditions. Is it a new hire? A board meeting? Hitting a growth milestone? A compliance deadline? Knowing the trigger lets you time your follow-up and your sales process correctly.
• "Is there a specific event or deadline coming up that makes this more pressing?" External forcing functions create urgency. Integrations breaking, contracts renewing, new funding rounds, executive changes - these are all moments when buyers move.
• "If we built exactly what you described, when would you want it available?" The answer tells you whether this is a now problem or a someday problem. "Yesterday" is a very different signal than "sometime next year."
• "Have you budgeted for a solution to this? Is there a specific budget cycle that affects when you could move?" B2B purchasing almost always has a budget calendar attached to it. If their fiscal year just ended, you might have perfect timing. If it just started, the money might already be allocated elsewhere.

No urgency usually means no deal, even if the interest is genuine. You want to understand what event or trigger creates urgency for them - a board meeting, a new quarter, a hiring surge, a competitor move.

Buying Behavior Questions (The Ones Nobody Asks)

This is where the interview gets uncomfortable for most founders - and that's exactly why you need it. You need to understand how buying decisions actually get made at this company, because getting someone excited about your idea and getting a check cut are two entirely different paths.

• "If something solved this tomorrow, who would need to sign off on the purchase?"
• "What does the buying process typically look like for a tool in this category - who else gets involved?"
• "Have you bought a tool to solve something similar in the past year? How did that decision happen?"
• "What would make you move on this quickly versus putting it on the backburner?"
• "Who else on your team would need to be involved or would push back?" In B2B, there's almost always someone who isn't in the room who can kill a deal. Legal, IT, the CFO - find out early who the hidden stakeholders are and what their concerns tend to be.
• "Have you ever championed a new tool or vendor internally? What made that process easy or hard?" This surfaces the internal political reality of buying. Some companies are fast movers. Some have 14-step procurement processes. You need to know which you're dealing with before you build your sales motion around their timeline.

The Willingness-to-Pay Questions (Do This Right or Skip It)

Willingness-to-pay is the most important data point in customer discovery and the one most founders collect worst. The classic mistake is asking directly: "How much would you pay for this?" That question produces useless answers. Everyone lowballs to avoid commitment, and nobody has a reference point.

Here's how to actually get at pricing signal:

• "What are you spending on this problem today, across all the tools, time, and workarounds?" The total cost of the status quo is your ceiling. If someone spends $3,000 a month duct-taping five tools together, and you solve the whole thing for a fraction of that, the math writes itself.
• "Have you tried to budget for fixing this before? What happened?"
• "If we built a solution and it cost around [anchor price], would that be in the right range or completely off?" Anchoring with a number and asking for a reaction is more useful than an open-ended price question. Watch their face as much as their words. If they wince, you're high. If they say "oh, is that all," you might be low.
• "What would you need to see to justify the cost internally?" This question gets you the ROI framing you'll need for every sales conversation. Let the customer write your business case for you.

The clearest willingness-to-pay signal is when someone tries to buy before you have a product. Even if it's just "can I be a beta user? Can I pay you now for early access?" - that's real intent, not conversation.

The One Question That Can Kill Your Idea (And Should)

Rob Fitzpatrick's The Mom Test makes a great point: you should always ask at least one question that has the potential to destroy your current idea. Most founders avoid this. That's exactly the wrong move.

Try: "Is there any reason you'd never switch away from what you're using now, no matter how much better something else was?"

Lock-in, switching costs, internal politics, compliance requirements - any of these can kill your market before you spend a dollar on development. Better to find out in an interview than after launch.

A few more idea-killers worth asking:

• "Is there something we haven't talked about that you think I should know?" This catch-all at the end of every interview often produces the most honest moment of the whole conversation. People save the real stuff for the end.
• "What would your boss say if you told them you were evaluating a new solution for this?" The internal reaction test. If the boss would immediately ask for a 20-page security audit, that's important to know now.
• "Do you see this problem getting worse, better, or staying the same over the next year?" If the trend is improving without a new solution, your window might be closing. If it's getting worse, urgency is only going to increase.

Questions for B2B vs. Consumer Discovery

The framework above applies broadly, but there are meaningful differences between B2B and consumer discovery that affect which questions you lean on.

For B2B discovery, the buying behavior and budget questions are non-negotiable. A single enthusiastic individual contributor who loves your idea but has no budget authority is not a validated customer - they're a cheerleader. You need to get upstream to whoever controls the budget and map the actual decision path. You also need to confirm company size and segment fit. A VP of Sales at a 15-person company and a VP of Sales at a 5,000-person company have completely different problems, budgets, workflows, and expectations, even if they have the same title.

For consumer discovery, the questions shift toward emotion, habit, and identity. Consumers don't fill out procurement forms or submit security questionnaires. But they do have deeply ingrained habits, social influences, and identity attachments that affect what they'll actually adopt. Ask more about what they've tried and abandoned, what they tell their friends about this problem, and whether they've ever paid for anything in this category before. "Have you ever paid for anything that tried to solve this?" is often more revealing than any hypothetical pricing question.

Building a Discovery Hypothesis Before You Start

One pattern I see in founders who get the most out of customer discovery: they start with written hypotheses, not just topics. Before the first interview, you should be able to write down: what you believe about who the customer is, what their primary struggle is, what they're currently using to cope, why that solution falls short, and what you think they'd be willing to pay.

The goal isn't to prove these right. The goal is to have a specific hypothesis to test so that contradictory information is easy to spot. When you go in without a hypothesis, you're just having conversations. When you go in with one, every interview generates signal.

Start with your biggest unknown - the assumption that, if wrong, kills your entire idea. One founder I know spent three months building a product for a problem he thought was urgent, only to discover in his first five customer conversations that his target buyers had just committed to a competing solution through a multi-year enterprise contract. If he'd started there, he'd have discovered it on day one instead of month three. Always test your highest-risk assumption first.

How to Find People to Interview

This question comes up constantly: where do I find people willing to do these interviews? The answer depends on your target market.

For B2B, LinkedIn outreach works well if you keep it short and frame it as research, not a sales call. Cold email is even better - a two-line note asking for 15 minutes of feedback from someone in a specific role converts better than most people expect. I've covered the cold email side extensively in the SaaS AI Ideas Pack - the research-framing approach applies directly to discovery interviews too.

If you need to build a list of people to reach out to quickly, this B2B lead database lets you filter by job title, industry, company size, and seniority so you're talking to actual decision-makers, not proxies. That's the difference between 20 interviews that teach you something and 20 interviews with people who have no budget authority.

If you need to find the direct email address for a specific person you've identified, an email finding tool can surface a verified contact quickly so your outreach actually reaches the right inbox. There's no point doing great discovery if your emails are bouncing before anyone reads them.

For consumer products, start with people you already know who fit the profile. Then expand to online communities - Reddit, Facebook groups, Slack groups, Discord servers - where your target customer already spends time. A direct message in the right subreddit asking for a quick call gets more responses than most people realize. The key is showing up where your potential customer already is, not asking them to come to you.

For local business products or services, scraping Google Maps for businesses in a specific vertical and geography gives you a fast list of real operators to cold email. ScraperCity's Maps scraper handles this if local businesses are your target segment.

How Many Interviews Do You Actually Need?

The honest answer: more than you think, fewer than you fear.

The research suggests a reliable read on market need takes 10-20 interviews with representative strangers - not three, and not necessarily 100. The key word is "representative." Ten conversations with the wrong people in the wrong segment teaches you almost nothing. Ten tight conversations with real buyers in your exact target segment can change everything.

Here's how I'd stage it:

• First 5 interviews: Pure exploration. You're not looking for patterns yet - you're learning what questions you should have been asking. After every interview, update your question set. You'll be surprised how different your approach is by interview five than it was at interview one.
• Interviews 6-15: Pattern-seeking. This is where you start listening for repetition. The same phrase from three separate people is a signal. The same frustration cited by five different people is a finding.
• Interviews 15+: Confirmation and edge cases. You're mostly validating what you already found, and you're testing whether the pattern holds in different sub-segments or company sizes. Stop when you're hearing nothing new.

If you're building something with significant capital investment attached - raising a round, hiring a team, committing six months of development - the number should be higher. Structural problem interviews with representative strangers, focusing on their problems, current solutions, and priorities, provides the evidence base needed to determine whether a real, painful, underserved need exists. There's no shortcut here that doesn't have a cost attached to it later.

How to Synthesize What You Hear

Don't try to remember your interviews. Take notes in real-time or record them (with permission). After each interview, write down the three most surprising things you heard. Not what confirmed your thesis - what surprised you.

After 10-15 interviews, patterns emerge. You're looking for phrases that repeat. When three different people describe a problem the same way, you've found your marketing copy. When five people cite the same frustration with a competitor, you've found your positioning. When nobody can name what they'd pay to fix this, you've found a problem that isn't a business.

A practical synthesis framework that works well:

• Create a simple spreadsheet with one row per interview and columns for: role, company size, problem described, current solution, frustration named, urgency level (1-10), willingness-to-pay signal, and key quotes.
• After each interview, fill in the row while it's fresh. Don't wait.
• After 10+ interviews, sort by urgency and willingness-to-pay. The people with both high urgency and clear budget signals are your early adopters. Build your initial ICP around what they have in common.
• Pull out every direct quote that describes the problem. These become your landing page copy, your email subject lines, and your sales script. Customers who describe the problem in their own words will recognize themselves instantly when they see that language reflected back at them.

If you're validating a SaaS idea specifically, the Business Idea Roaster is a useful complement - it stress-tests the concept before you go deep on interviews.

What Good Signal Looks Like vs. Polite Noise

This is the part most discovery guides skip. They tell you what to ask but not how to tell the difference between real signal and someone being polite.

Here's what genuine signal looks like in a customer discovery interview:

• They lean forward (physically or metaphorically) when you hit the right topic. Energy goes up. They start talking faster and with more detail.
• They reference a specific time this cost them money, a customer, a deal, or a night's sleep.
• They offer to make an introduction: "You should talk to my colleague Sarah - she deals with this even more than I do."
• They ask how they can stay in the loop when you're ready to launch. Nobody does this out of politeness.
• They push back on your framing: "Actually the biggest issue isn't X, it's Y." Contradiction is a gift - it means they're engaging with the reality, not the polite version.

And here's what polite noise looks like - the stuff that feels like a good interview but isn't:

• "That sounds really interesting." Full stop. No specifics, no follow-up questions from them, no concrete problem named.
• "I'd probably use something like that." The word "probably" in a discovery interview is nearly always a red flag.
• "I can definitely see the market for this." They're commenting on your business idea, not their own problem. The subject has shifted from them to you.
• Lots of enthusiasm, no specifics. If someone can't name the last time the problem hurt them in a concrete way, they're being supportive, not honest.

Real customer discovery is uncomfortable precisely because real signal is harder to earn than encouragement. If every interview feels great and you leave feeling validated, be suspicious. That's often what getting flattered looks like.

Common Mistakes to Avoid

• Pitching during the interview. The second you describe your solution, the honest conversation is over. The moment you introduce your product, the customer stops describing the world as it is and starts reacting to your idea. Save it for the end, or for a follow-up call.
• Asking hypothetical questions. "Would you use a tool that..." is not customer discovery. "What have you actually tried?" is. Past behavior predicts future behavior. Hypotheticals predict nothing. The moment you catch yourself asking "would you" instead of "did you" or "what happened when" - reset.
• Talking to people who want to be nice to you. Friends, family, and warm contacts who respect you will tell you what you want to hear. That's why cold outreach for discovery interviews is actually an advantage - strangers have no incentive to flatter you.
• Leading questions. "Don't you find it frustrating when X happens?" is not a discovery question - it's a confirmation request with a polite "no" built out of reach. Questions that suggest the answer contaminate the data. If your question has an obvious "right" answer embedded in it, rewrite it.
• Stopping at 3-4 interviews. You need enough conversations to see patterns, not just anecdotes. Ten is a minimum. Twenty is better if you're building something with significant investment attached. The data suggests pivoting only when you've made at least 40 discovery conversations with consistent negative feedback on your core value proposition - not after three mixed ones.
• Ignoring what you didn't expect. The signal is almost always in the surprises, not the confirmations. If you're only writing down things that match your thesis, you're not doing discovery - you're doing confirmation bias with extra steps.
• Only talking to one level of the organization. In B2B especially, you need conversations at multiple levels. The individual contributor tells you about the daily pain. The manager tells you about the team impact and the budget process. The executive tells you about strategic priority and buying authority. You need all three perspectives before you can build a complete picture.

Turning Discovery Into an ICP and Positioning

Customer discovery isn't just about validating whether a problem exists. It's about generating the raw material for everything that comes after: your Ideal Customer Profile, your positioning, your messaging, your pricing logic, and your sales process.

The ICP that comes from good discovery is specific. Not "marketing managers at mid-sized companies" - but "marketing managers at B2B SaaS companies with 50-200 employees who manage outbound sequences manually and are frustrated by their current tool's limited reporting, specifically around sequence-level attribution." That level of specificity only comes from real conversations. No amount of desk research gets you there.

Your positioning comes from the pattern of frustrations you heard. If seven out of twelve people you interviewed cited the same limitation in the current solution, that limitation is your positioning anchor. You don't need to manufacture differentiation - your customers already told you what's missing.

Your messaging comes from their exact words. The phrases your customers use to describe the problem - pulled verbatim from interview notes - are almost always better marketing copy than anything you'd write from scratch. People recognize language that sounds like themselves.

I dig into the practical side of this - building the right ICP, running outbound research, and turning discovery data into positioning - inside Galadon Gold with the rest of the coaching community.

What Good Customer Discovery Actually Produces

When you run this process well, you walk away with three things that most founders don't have before they build: a specific description of the customer (not a vague persona, but a real segment with a real job title and real context), a precise articulation of the problem in the customer's own language, and evidence of willingness to pay - ideally a few people who've said they'd buy if you built it.

That last part matters more than anything else. If nobody you interview offers to give you their email, introduce you to a colleague, or ask to be notified when you launch, the interviews weren't as positive as they felt. People vote with their actions, not their words.

The startups that survive long enough to find product-market fit are distinguished not by luck but by rigorous customer discovery, capital discipline, and the ability to recognize and respond to market feedback quickly. Discovery is what gives you the feedback worth responding to.

For daily frameworks and ideas on business validation and growth, the Daily Ideas Newsletter is worth having in your inbox.

Customer discovery is the unsexy work that separates companies that ship the right thing from companies that ship fast and pivot constantly. Do it right and you'll have a roadmap. Skip it and you'll have a launch plan built on guesses.