What we measure
- Creator overlap. How many independent creators recommend the tool in the rolling window. Overlap across distinct channels is the highest-signal metric — one creator with a million views counts less than five creators with ten thousand each.
- Recommendation frequency. How often each creator returns to the tool. A tool mentioned once is noise; a tool mentioned across multiple videos is conviction.
- Sentiment polarity. Whether the recommendation is positive, mixed, or comparative — counted with directional context against the alternative tool.
- Recency weighting. Videos in the last 90 days count fully. Older videos decay so legacy recommendations don't pin a stale ranking.
The creator source
YouTube Consensus
Framework onlyAggregated creator recommendations — counts creator overlap, recency, and sentiment, not raw view counts.
How the Creator Consensus Score is calculated
- Identify the creator universe — channels with sustained founder/AI-tool coverage and a minimum subscriber threshold.
- Extract tool mentions from titles, descriptions and transcripts, deduplicated per channel per tool per window.
- Score per tool = overlap × frequency × sentiment × recency, normalized to 0–100 per category.
- Surface narrative — the most common praise points and the most common criticisms become the Creator Narrative block on the tool profile.
Guardrails
YouTube Consensus is framework-only today. Until ingestion is live, Creator Consensus Scores, creator winners, and Creator Narrative blocks render as "Signal coming online" and never contribute to Consensus Scores, FutureFounder Score™, awards, reports, or insight rankings.
When live, we publish the creator universe in the open. Sponsored video segments are excluded and flagged. See /sources for the current data state.
Last updated January 2026