Product

What the 0–100 AI Fit Score actually measures

2026-07-08 · 4 min read

By · Reviewed 2026-07-08

It is not a quality grade and not a guarantee — it is a relevance ranking built from evidence. Understanding what moves it makes you faster.

Relevance to your brief, weighted by evidence

The score compares what your brief asked for — market, product family, channel role, exclusions — against what the agent verified on the company’s real pages. Verified matches move the score; unverifiable claims do not.

Why two similar companies score differently

Usually: evidence depth. A company with a matching category tree, an importer registration, and a purchasing contact page gives the model more verified surface than one with a single-page site — even if both are “relevant”.

How to use it well

Sort by score, then read evidence top-down. Contact the top band this week, batch-research the middle band, and refine your brief if the whole list scores low — a low-scoring list is feedback about the brief, not the market.

Key takeaways

  • The score ranks relevance to your brief — it is not a company quality grade.
  • Verified evidence moves the score; marketing claims do not.
  • Low scores across the whole list mean the brief needs work, not the market.

More from the TradeOcto blog

Explore TradeOcto