Continuous Discovery
A sustained cadence of customer contact and small research activities that informs product decisions alongside ongoing delivery work.
Download Markdown (.md)Soroush’s Product Wiki · https://wiki.soroush.uk/topic/continuous-discovery/
01What it means
Continuous discovery is the habit of integrating frequent customer touchpoints and rapid learning activities into a product team’s regular work. Teresa Torres describes it as, at minimum, weekly customer contact by the team building the product, conducted in pursuit of a desired outcome.[1][2]
- 01Current decisionOpportunity, assumption, or tradeoff needing evidence
- 02Small research activityInterview, observation, prototype, data, or experiment
- 03Synthesize evidencePatterns, contradictions, uncertainty, and limits
- 04Choose + deliverAdapt direction or test a bounded solution
- 05Observe outcomesCustomer and operational effects after release
New outcomes expose the next decision and evidence gap
Continuous discovery is a decision habit, not a quota of interviews; evidence quality, participant diversity, ethics, and strategic coherence still matter.
02Why it matters
Customer knowledge decays as markets, products, and behavior change. A steady cadence reduces reliance on occasional research projects, shortens the distance between a question and evidence, and allows the cross-functional team to adjust decisions before assumptions harden into expensive commitments.
03When to use it
Use it for products with ongoing decisions, reachable customers, and a stable cross-functional team. It is especially useful when teams ship frequently or operate in changing domains. Adapt the cadence when access is constrained or participant burden is a concern.
04How it works
- Set an outcome that gives discovery a decision-making focus.
- Create ethical, repeatable recruiting channels rather than scrambling for participants.
- Hold frequent customer touchpoints with product, design, and engineering involved.
- Capture opportunities and evidence in a shared, revisable model.
- Test solution assumptions in small increments and compare alternatives.
- Connect learning to delivery choices and track outcomes after release.
05A practical example
A collaboration-tool trio schedules two customer conversations each week. While pursuing faster project setup, it hears repeated confusion about ownership, maps that opportunity, tests three onboarding explanations, and ships the smallest successful version while continuing to monitor activation data.
06Where it breaks
Weekly contact is not automatically good research. Convenience samples, repeated exposure, weak questions, and shallow synthesis can amplify bias. The practice also requires recruiting operations and team capacity, and it should not pressure vulnerable customers or replace specialist research.
Common mistakes
- Counting any customer call as discovery regardless of the decision it informs.
- Delegating all contact and synthesis to one researcher or designer.
- Maintaining the cadence while failing to change priorities when evidence shifts.
07Sources
- Continuous Discovery: Definition and OverviewProduct Talk · Creator-authored glossaryContinuous Discovery: Definition and Overview
- Continuous DiscoverySilicon Valley Product Group · Practitioner articleContinuous Discovery