How the weekly Momentum Score is calculated — and what data feeds it.
Every Monday we refresh a 0–100 Momentum Score for each company by blending four independent public signals. Each signal is log-normalized across the full field so a single viral spike can't dominate, then weighted and combined. A short-term trend bonus (±20) rewards companies whose Wikipedia interest is accelerating week-over-week. Final ranks (1 through N) are produced by sorting on the composite score.
When a data source is temporarily unavailable, its weight is redistributed across the remaining live signals — the score stays comparable across weeks.
Daily English Wikipedia article views for each company, rolling 7-day window vs the prior 7-day window.
Baseline of general public awareness. Spikes correlate with news, launches, controversy, and recruiting pushes.
https://wikimedia.org/api/rest_v1/Count of videos uploaded in the last 7 days that mention the company by name (YouTube Data API v3 search).
MLM lives on YouTube — opportunity videos, comp-plan reviews, upline trainings. Velocity here is a leading indicator of rep activity.
https://developers.google.com/youtube/v3Number of currently-running ads in the U.S. mentioning the company name (Meta Ad Library API).
Active ad spend on recruiting creative is the clearest public signal that a company is pushing growth right now.
https://www.facebook.com/ads/library/api/Posts in the last 7 days containing the company name as an exact phrase (reddit.com/search.json).
Captures unfiltered community chatter — both fans and critics. Sentiment is mixed by design.
https://www.reddit.com/dev/api/We'd rather show fewer, defensible signals than pretend to have data we don't. These are on the roadmap when we can do them properly: