Skip to main content

The market always
knows more than you

No discovery, however deep, replaces real contact with the market.

Statement​

The signals generated by real usage are higher-quality data than any prior hypothesis. The system must be designed to listen to them and act on them. The market doesn't lie, doesn't speculate, and has no confirmation biases.

Why it matters​

The most exhaustive discovery is still a hypothesis until the market validates or invalidates it. Problem-Driven AI isn't about guessing better, but about creating a system that listens better.

Market signals are first-class data: usage metrics, direct feedback, churn patterns, recurring requests. Ignoring them in favor of internal hypotheses is an act of methodological arrogance.

Practical implications​

  • Design for observation: every deliverable should include mechanisms to capture market signals.
  • Prioritize real usage data over internal opinions or theoretical research.
  • Establish listening cycles: don't wait for the market to scream. Design channels to hear its whispers.
Anti-pattern: The Eternal Discovery

What it is. Staying in the research phase indefinitely, refining hypotheses without exposing them to the market. Research becomes a refuge from the uncertainty of real user contact.

How to detect it. The team has been in discovery for multiple cycles without producing a testable artifact. Every finding generates new questions but never a clear enough answer to move forward. The phrase "we need more data" is used to delay decisions, not to improve them. No Exit Criteria for the discovery phase have been defined.

How to prevent it. Define Exit Criteria for every discovery cycle before it begins. Set a maximum number of iterations per discovery phase. Force the team to produce a testable hypothesis after each cycle, even if imperfect. The market will refine what research cannot.

Connections​