If you ask someone whether your startup idea sounds good, they will almost always smile and say yes. Friends do it to encourage you, colleagues do it to be polite, and strangers on social platforms do it because saying something is cool requires zero effort and costs zero dollars.
Relying on that encouragement is the most reliable way to waste four months building an application nobody opens twice.
Customer discovery is not about pitching a vision or validating your excitement. It is forensic examination of past behavior. People are remarkably bad at predicting what they will pay for in the future, but they are completely honest when recounting how they solved a problem last Tuesday and how much money left their company bank account to fix it.
Here is how I structure 15-minute discovery interviews as a solo operator, the specific question templates that avoid confirmation bias, and the exact signals that tell you to drop a premise before writing code.
The Core Rule: Ban the Future Tense
Whenever a conversation slips into conditional phrasing—“Would you use a tool that automatically synced invoices?”—you have lost the interview. The respondent enters an imaginary universe where every workflow is frictionless and software subscriptions cost nothing.
The rule I enforce is simple: ban the future tense and ban the hypothetical. You do not talk about your concept, your features, or your proposed pricing tier.
Instead, anchor the discussion to an actual event:
- Weak (Hypothetical): “How much would you pay for an automated invoice reconciliation dashboard?”
- Actionable (Historical): “When was the last time you reconciled client invoices? Walk me through what you did step by step.”
When you force someone to describe reality, the friction becomes visible immediately. You see whether they spent four hours wrestling with an Excel sheet or whether they shrugged, sent three emails, and moved on. If they solved the problem with an ugly spreadsheet and have tolerated that routine for two years without searching for a commercial alternative, their pain is not acute enough to sustain a standalone SaaS product.
4 Questions That Reveal Genuine Purchasing Intent
During a 15-minute call, I never ask more than four primary questions. Everything else is a brief follow-up clarifying specific mechanics.
1. “What is the hardest part about handling this workflow?”
This opens the floor without steering them toward your predetermined solution. Notice the phrasing: it does not ask what they dislike about their current software; it asks what is difficult about the operational workflow itself. If their primary complaint is an organizational or communication bottleneck rather than a tooling failure, software cannot solve it.
2. “Why was that particular instance so frustrating?”
You need to identify whether the cost is emotional annoyance or measurable business drag. Mild irritation does not unlock corporate credit cards. Lost revenue, compliance deadlines, security audits, and client refund disputes unlock credit cards.
3. “What have you actively tried to do to solve this?”
This is the single most diagnostic question in early validation. If the prospect answers, “We have just dealt with it,” disqualify the idea. Sustainable businesses are rarely built by convincing indifferent people to care about a problem for the first time. They are built by offering a superior, faster, or more focused alternative to people who are already spending time or money attempting to fix it.
Look for existing compromises:
- Have they strung together Zapier webhooks and Airtable bases?
- Did they hire a part-time virtual assistant to handle data entry?
- Are they paying for an enterprise tool where they only touch two features?
4. “What do you dislike about the workaround you just described?”
This establishes your product roadmap. Their answer tells you the precise operational failure of existing market solutions—whether it is exorbitant per-seat pricing, terrible mobile responsiveness, or complex API configurations.
How to Recruit 10 Participants Without Cold Spam
Founders frequently assume recruiting interview subjects requires massive advertising budgets or automated cold outreach spam. In practice, genuine discovery calls require low-volume, consultative outreach.
Search communities where practitioners congregate—industry-specific subreddits, professional Discord servers, Slack user groups, or LinkedIn search. Send a three-sentence inquiry that explicitly states you are not selling anything:
“Hi [Name], I am researching how independent design studios handle client milestone approvals and scope disputes. Not pitching or selling any software. Would you be open to a 10-minute chat about your current workflow next week? Happy to share our summarized notes afterward.”
People enjoy talking about their operational headaches when they are confident they will not be subjected to a high-pressure sales pitch at minute fourteen.
If an interviewee asks what you are building, resist the urge to deliver a demo. Tell them you are still assessing whether the problem is worth solving. If their frustration was severe, they will ask you to notify them whenever you ship a solution—and that is your first verified pre-launch lead.
For more frameworks on stress-testing early concepts, read our breakdown on the smoke test playbook for selling before building and our field guide on 5 cold email signals that tell you to kill an idea in 7 days.