Discovery · Evolved from Continuous Discovery (Torres)
Evidence-Weighted Continuous Discovery
Weekly touchpoints with customers, structured around opportunities and assumptions. Modernized with explicit evidence provenance, AI-assisted synthesis, and human validation of every AI-generated claim.
Executive Guidance
How to hold this framework as a leader
Continuous discovery is the operating cadence that keeps decisions tethered to customer evidence. In enterprise environments the constraint is not motivation but access — customer time is gated and expensive. The discipline is therefore about rationing access into a repeatable weekly rhythm that produces artifacts a leader can rely on.
AI has meaningfully changed the interior of this practice: synthesis, transcription, and clustering are now fast and cheap. That shift makes provenance the executive concern. Every insight that influences a decision must carry a citation to a real interview, or the roadmap begins to inherit AI's hallucinations at scale.
When to Apply
- Product teams shipping regularly but uncertain about impact
- Enterprise contexts where customer access is gated and precious
- Any environment where AI synthesis of research is being introduced
AI-Era Notes
AI can accelerate transcription, clustering, and pattern surfacing — it cannot replace direct customer contact. Enforce source citations on every AI-derived insight and require human review before an insight influences prioritization.
Key Trade-off
AI synthesis accelerates throughput but risks laundering unattributed claims into roadmap decisions.
Operating Sequence
The order in which to install the framework
- 01Commit the team to a weekly touchpoint cadence and defend it as a leadership expectation.
- 02Define the decision each week's discovery is meant to inform.
- 03Run structured interviews with a common opportunity/assumption schema.
- 04Use AI to accelerate transcription and clustering; require human validation before roadmap use.
- 05Publish a weekly discovery digest with citations, updated opportunities, and open assumptions.
Key Artifacts
The documents this framework produces
Interview Guide
Consistent structure and open prompts across weekly sessions.
Owner: Product + Research
Opportunity Backlog
Living list of validated customer opportunities tied to outcomes.
Owner: Product
Assumption Register
Explicit product bets, their status, and evidence status.
Owner: Product
Discovery Digest
Weekly summary read by leadership with cited insights.
Owner: Product
Operating Checklist
What "good" looks like when installed
Cadence
- At least one customer touchpoint per week per team.
- Discovery has a named decision it is informing this quarter.
- Weekly digest is distributed to leadership and stored durably.
Evidence Discipline
- Every insight in the digest cites a specific interview or observation.
- AI-generated syntheses are reviewed by a human before circulation.
- Verbatim quotes and interview provenance survive summarization.
Assumptions
- Assumptions are explicit, testable, and owned by a named person.
- Assumption status is updated in the register within a week of new evidence.
- Critical assumptions have a defined next test.
Common Antipatterns
- Replacing interviews with AI persona simulation
- Losing verbatim quotes and provenance in summarization
- Running discovery without a decision it will inform
Boardroom Questions
- What did we learn from customers last month that changed a roadmap decision?
- Which of our largest bets rest on assumptions we have not yet tested?
- Where has AI-assisted synthesis introduced claims we cannot source?