How ranking works
Votes rank launches. Buyer intent ranks discovery.
Live fair signal
score = Σ(voteWeight) × commentBoost × recencyscore = sum(voteWeight) × commentBoost × recencyFactor × intentFactor- voteWeight = 1 if account ≥ 7 days; else 0.25
- commentBoost = 1.15 with structured feedback; else 1.0
- recencyFactor soft-decays so day-1 pile-ons cannot freeze the board
- intentFactor = 1 + 0.1 × min(5, useSignalScore). Fits, Would try, and Would pay signals gently boost live week without freezing cold starts.
evergreen = log(1 + weightedVotes) + feedbackQuality + useSignalScore + needMatchBoost + halfLife(90d)- useSignalScore: Would pay (1.2) > Would try (0.7) > Fits (0.35), log-compressed
- needMatchBoost: Only seeker-confirmed interested matches (0.35 each, cap 2). Automated suggestions alone never inflate rank.
Products stay findable in Browse and match open needs.
karma = min(votesGiven) + min(feedbackGiven) + min(received)- Karma tracks helpful participation. It does not change product launch or evergreen scores. It is also not the main maker goal. Look for need matches and buyer signals instead.
- Structured feedback (3 pts) beats empty congrats (1). Caps: votes given 50, feedback 150, received 100.
- Opaque Featured vs All tabs
- Do-or-die 24-hour windows
- Treating brand-new accounts like established ones
- Congrats-only comment culture
- Ranking bought with Launch Pass or Premium
