Market Volume covers the products with observed sales. Relevance covers the rest — new releases and the long tail — by what the market shows: category position, ratings, bestseller signal, retailer coverage. AI-validated against real sales when they appear.
Half the catalog never registers a sale this month. New releases. The long tail. The products everyone is guessing about.
Market Volume tells you what sells. Relevance tells you what's about to. Visibility signals — category position, customer ratings, bestseller signal, retailer coverage — score every product before sales data can reach it. AI-validated against real sales when it finally does.
Average category ranking across direct-stock listings on key retailers, last thirty days. Worth fifty-seven and a half points.
Stars scaled by review volume across all listings. Forty-five hundred reviews earns full weight; below that, proportional.
Position on marketplace bestseller lists for the region. Five points for rank one, decaying to zero by rank one hundred.
Three of five key retailers carry it directly? Multiplier is point six. Coverage is half the answer.
Bestsellers don't decay. Everything else does, on a curve. Old products lose weight automatically.
The AI Relevance Analyser then runs statistical analysis across OpenAI, Anthropic, and Google models — computing sales concentration, sales multiplier, and marketplace correlation against your real catalog. The score isn't a black box. It's a derivation you can defend in a board meeting.
We run Relevance on one of your categories. Within ten seconds you see what scores high — and how it correlates with the products that already have sales data.