---
name: product-discovery
description: Examine customer or market evidence to frame an unmet need, competing explanations, and the next useful product learning step.
---

# Product Discovery

Help determine which customer problem merits product attention. Establish the
decision, affected people and supplied constraints. Inspect available artifacts
for factual answers first. Ask focused questions about product choices or
material evidence gaps; do not require an interview when context already
supports useful work.

When asked to deepen the work, or a consequential framing depends on a summary
or an untested research method, use [research depth](../_shared/references/research-depth.md).
Improve the evidence and method for the same decision; routine settled work
needs no additional discovery exercise.

Inventory evidence before interpreting it. Preserve origin, date or period,
collection method, sample and selection limits. Separate observed behavior,
customer statements, interpretations and hypotheses. Keep conflicting and
negative evidence visible. Do not infer prevalence, demand, market size,
revenue or causation from a handful of examples. Include customer identifiers
only when needed for the authorized decision.

Use the domain's existing terms and clarify ambiguous concepts from sources.
Describe the situation, user's goal, current alternatives, workarounds and
friction before proposing features. Follow a concrete episode through the
steps where the problem appears; distinguish stated preferences from actual
behavior and observed consequences.

When validating an idea before commitment, read
[decision pressure-testing](../_shared/references/decision-pressure-testing.md)
to identify the consequential assumption, credible alternatives and evidence
that could overturn the direction.

When drafting research questions, ask about past events, choices and outcomes
without suggesting the desired answer. Seek disconfirming cases and alternate
explanations. A question about willingness to use an imagined feature does not
establish adoption. Drafting research does not authorize recruiting or contacting
participants.

Frame an opportunity around a user and unmet need. State what is known,
inferred and still uncertain, and what evidence would change the framing.
Propose the smallest useful observation or experiment when uncertainty prevents
a choice. Use `product-experiments` to define that test; use
`product-prioritization` to compare already framed opportunities.

Finish with a concise brief: decision and scope, evidence and sample limits,
opportunity, competing explanations, and next learning step. Preserve open
product decisions rather than manufacturing confidence. Discovery does not
commit implementation, delivery dates, outreach or provider actions; these stay
within the user's task authority.