Reference
If it cannot be shown, it does not count.
Evidence Definitions are the canonical rules behind every ProductBooks verdict.
System Flow
Problem → Product → Model
Problem
Problem–Solution Fit
If the problem is weak, every signal above it misleads.
Product
Product–Market Fit
If users do not return, you do not have product–market fit.
Business Model
Business Model Fit
If margins break at scale, the model fails.
If one breaks, everything above misleads
- Product signals are invalid without a real problem.
- Business signals are invalid without real product pull.
Evidence & Verdicts
Strict inputs only. Missing proof does not get the benefit of the doubt.
What counts as evidence
- Source material from real users or traceable systems.
- Observable behaviour, stated preference, or measurable outcome.
- Evidence that matches the stage being assessed.
What scores zero
- Paraphrased summaries with no source proof.
- Team belief, optimism, or interpretation.
- Forecasts used as current evidence.
How ProductBooks applies the rules
The evaluator scores what is present, not what the team meant.
Direct evidence carries more weight
- Your own user interviews.
- Your own analytics tied to real usage.
- Your own payment or transaction records.
Indirect evidence is weaker
- Third-party reports.
- Competitor reviews.
- Secondary research without direct user proof.
What the rules are protecting
The system exists to stop drift, not to win arguments.
The reference prevents
- Score inflation through interpretation.
- Confirmation bias disguised as validation.
- Moving stages without the required proof.
What to do next
Use the reference, then return to the stage you are trying to clear.
Evidence Definitions explain the rules. The evaluator still decides whether the current proof passes the stage.
Go to Problem–Solution FitRun the stage once the evidence is ready to be shown.
ProductBooks does not need a stronger story. It needs stronger proof.