ChatGPT does not simply find the best store. It constructs a shortlist, evaluates what it knows about each brand, and decides which recommendations it can make with confidence.
This is what happens between a buyer's prompt and a brand recommendation. Every step here runs in seconds, and most store owners never see any of it.
Every candidate brand passes through the same four checks. A weak result at any layer can remove a brand from the shortlist, no matter how strong the others are.
This is exactly why Prompt Visibility has to be measured prompt by prompt, and why Recommendation Share is measured across a whole set of buyer prompts, not a single query.
There is no single "ChatGPT score" for a brand made of fixed percentages. What ChatGPT can know about you comes from three places, and all three feed the same evaluation.
Most stores that check their own name in ChatGPT are only testing whether they are known. See Unverified Visibility for what happens in that gap between known and recommendable.
what you sell + who it's for + use case + differentiator. Generic positioning matches nothing well.AI recommendations are mostly invisible to store owners. Your existing tools were not built to show them.
Run your store through the same intelligence layer we're measuring.
Illustrative example · single-site signal for atomfoundry.dev.
View full signals →