Problem interviews and the art of customer discovery

The Mom Test is a framework designed to help founders get honest data by asking about a person’s actual life and past behaviors rather than pitching a future idea. It is named after the concept that your mother loves you and will lie to protect your feelings if you ask for her opinion on an invention. By focusing on specific past actions—such as how someone recently solved a problem or what they spent money on—you prevent "contaminating" the data with polite, optimistic lies or "false positives."

Laddering is a systematic way of asking "why" to move from surface-level functional complaints to deep emotional and social motivations. By digging into the consequences of a problem—such as how a slow process affects a manager’s reputation or leads to lost accounts—you can determine if a product is a "vitamin" (nice to have) or a "painkiller" (essential). This helps identify "hair-on-fire" problems, which are high-magnitude issues that customers are desperate to solve regardless of the cost or aesthetic of the solution.

Solution Verification is an experimental phase where you validate a potential solution with the minimum amount of work possible. Unlike a sales pitch, which tries to convince a user to buy, verification uses "Smoke Tests" or "Concierge MVPs" to see if users actually want the outcome. Examples include manually performing a service before automating it or using a landing page with a "pricing" button to measure real "Willingness to Pay." The goal is to collect "Truth Data," such as email sign-ups or Letters of Intent, rather than just verbal agreements.

Kill Criteria are hard thresholds or benchmarks set before an experiment begins to protect founders from their own optimism and confirmation bias. Because entrepreneurs often suffer from the "Sunk Cost Fallacy," they may continue building a product even when data is negative. By deciding in advance that a project will be pivoted or killed if it doesn't meet a specific metric—like a 15% conversion rate or five pre-orders—founders can ensure they are making data-driven decisions rather than emotional ones.

In the modern landscape, startups can use a hybrid approach by employing AI-moderated interviews to gain "Quantitative Breadth" across hundreds of participants while maintaining 10 to 15 deep, personal "Problem Interviews." While AI is a powerful "Learning Accelerator" that can identify patterns and summaries, human conversations are still required to capture empathy, nuance, and the emotional "why" behind a customer's frustration. This balance allows a company to stay close to the customer's pain while moving at a high "Validation Velocity."