The loop restarts from Signal Collection after every launch.
Continuous improvement cycle
1
PRD
The starting point of each loop. Defines the product direction, key features, and success criteria. Updated in parallel during the Measure and Feedback phase so it stays current across cycles.
Loop anchor
2
Signal Collection
Collecting signals from external sources. Log new data into the Signal log as it arrives - not necessarily on a weekly schedule.
- Public reports and web
- Social listening - users
- Social listening - competitors
3
Insight and Hypothesis
Analyze signals, identify patterns, gaps, and opportunities. Log hypotheses into the Hypothesis log and track status: untested / testing / confirmed / rejected.
4
Recommendation v1
Initial strategic direction based on insights and hypotheses. Document the rationale clearly so the Founder can assess risk level. This is the input for the risk decision gate.
5
Risk Assessment
The Founder decides the risk level and determines the next step in the flow.
- Low risk - proceed directly to Build and Launch
- High risk - enter the Validation loop before committing
5b - Validation (High risk only). Runs only when the Founder assesses High risk. Each validation round creates a new document. Includes User Research (interviews + surveys), Market Research (market and competitor analysis), and Recommendation v2 (updated direction after validation). Returns to Risk Assessment gate.
6
Build and Launch
Execute based on the finalized recommendation - either Recommendation v1 (low risk) or Recommendation v2 (after validation). Build, test, and ship the product increment.
7
Measure and Feedback
Measure outcomes after launch. Collect user feedback and product metrics. PRD is updated in parallel to reflect what was built and what was learned. The loop then returns to Signal Collection.
Returns to step 2