Skip to content

nWave-ai/nWave / nWave/skills/nw-pdr-review-criteria

nw-pdr-review-criteria

Verified

Evidence quality validation and decision gate criteria for product discovery reviews

Review Criteria -- Product Discovery Review

Evidence Quality Validation

Past Behavior Indicators (Good)

"Tell me about the last time..." | "When did you last..." | "What happened when..." | "Walk me through how you..." | "What did you try..." | "How much have you spent on..." Specific dates, dollar amounts, named tools, concrete examples, emotional frustration language

Future Intent Red Flags (Reject)

"Would you use/pay/like..." | "Do you think..." | "Imagine if..." | "What if we..." Flag and reject if >20% of evidence is future-intent.

Validation Thresholds

  • Past behavior ratio: >80% (reject if fail, require re-interview)
  • Specific examples: min 3 concrete per finding (warn if fail)
  • Customer language: quotes in customer words, not paraphrased (warn if fail)

Sample Size Minimums

PhaseMinimumHigh ConfidenceNotes
1: Problem510Interviews required
2: Opportunity1020Quantitative supplements, not replaces
3: Solution5 per iteration3 iterations maxBefore decision
4: Viability5--Stakeholder review required

Decision rule: min 5 actual, consistent signals supporting the decision. Block decisions on fewer; do not infer missing customers or discard negative evidence to satisfy a threshold.

Decision Gate Criteria

G1: Problem to Opportunity

Proceed: 5+ interviews, 5+ participants confirm pain, >60% confirmation across all Phase 1 interviews, willingness to pay supported by past spending or observed commitment | Pivot: differs from expected | Kill: <20% confirm Checks: verify both confirmation count and strict percentage from actual interview evidence, preserve sources and disconfirming observations, customer words, 3+ examples. 3/5 fails both requirements; 4/5 fails the count; 5/8 passes both count and percentage.

G2: Opportunity to Solution

Proceed: top 2-3 score >8 (0-20) | Pivot: new opportunities | Kill: all low-value Checks: OST complete with 5+, scores correct (Importance + Max(0, Importance - Satisfaction)), top >8/20

G3: Solution to Viability

Proceed: >80% task completion, usability validated | Pivot: needs refinement | Kill: fundamental blocks Checks: 5+ users/iteration, >80% completion, core flow usable, value validated

G4: Viability to Build

Proceed: 4 risks addressed, model validated | Pivot: adjustment needed | Kill: no viable model Checks: Lean Canvas complete, all risks green/yellow, stakeholder sign-off

Bias Types

Confirmation Bias (critical)

Signals: only positive quotes | skeptics not interviewed | disconfirming evidence dismissed | same questions for "right" answers Fix: include skeptics, actively seek disconfirming evidence

Selection Bias (high)

Signals: all existing customers | no churned/non-adopters | lacks diversity | single-enthusiast referral chain Fix: random/diverse selection, include skeptics and non-users

Discovery Theater (critical)

Signals: conclusion decided before research | findings match hypothesis perfectly | no surprises | idea-in = idea-shipped Fix: track idea evolution, expect 50%+ ideas to change

Sample Size Problem (high)

Signals: major decisions on 2-3 interviews | single quote as "validation" | pivot on one signal Fix: min 5 interviews per segment, 5+ signals for decisions

Anti-Patterns

Interview Anti-Patterns

PatternDetectionBadGoodSeverity
Leading questionsSuggests desired answer"Don't you think this would save time?""Tell me about the last time you tried to save time on this"high
Future-intentHypothetical behavior"Would you use this feature?""What have you tried to solve this problem?"critical
Compliments as validationAccepting "that's cool""They loved the idea!""They committed to follow-up and referral"high
Talking > listening>30% interviewer talkLong questions, short responsesOpen questions, extended answersmedium

Process Anti-Patterns

PatternDetectionSeverity
Skipping to solutionsSolution before problem validatedcritical
Demographic segmentationSegments by demographics not jobsmedium
Building before testingCode before Phase 3critical

Strategic Anti-Patterns

PatternDetectionSeverity
Premature pivotingDirection change on 1-2 signals (need 5+)high
Solution loveDefending despite evidence, dismissing criticshigh
Sole source of truthOnly quant OR qual, not bothmedium

Pre-Approval Checklist

Evidence Quality

  • Past behavior ratio >80% | [ ] No critical future-intent | [ ] Customer language preserved

Sample Sizes

  • Phase 1: 5+ | [ ] Phase 2: 10+ | [ ] Phase 3: 5+/iteration | [ ] Phase 4: 5+ with stakeholders

Decision Gates

  • All gates properly evaluated | [ ] Criteria documented with evidence | [ ] Decision justified

Bias Check

  • No confirmation bias | [ ] No selection bias | [ ] No discovery theater | [ ] Sample size adequate

Anti-Patterns

  • No critical interview anti-patterns | [ ] No critical process anti-patterns | [ ] No critical strategic anti-patterns

Approval Decision

First matching outcome wins:

  • Rejected: supplied evidence fails a criterion, or any critical/high issue. Structured rejection with remediation. Blocks handoff.
  • Indeterminate: nothing fails on supplied evidence, but a criterion's needed evidence is absent or unreadable. Name the missing evidence and its supplier. Blocks handoff; never approval.
  • Conditionally approved: every criterion judged, minor issues only (no critical/high). Approval with recommendations.
  • Approved: every criterion judged and passing, no critical issues. Formal handoff to product-owner.
receiving-code-reviewobra/superpowersUse when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementationStars: 322kUpdated Claude CodeCodexVerifiedInstall
requesting-code-reviewobra/superpowersUse when completing tasks, implementing major features, or before merging to verify work meets requirementsStars: 322kUpdated Claude CodeCodexVerifiedInstall
using-git-worktreesobra/superpowersUse when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallbackStars: 322kUpdated Claude CodeCodexVerifiedInstall
verification-before-completionobra/superpowersUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions alwaysStars: 322kUpdated Claude CodeCodexVerifiedInstall