GetSeenInAICheck my brand

Methodology

How we score this

Last updated 15 Jul 2026

The score answers one question: when a shopper asks ChatGPT for a product like yours, does it bring you up, and how does it talk about you? Here is exactly how we measure that, so you can trust the number and explain it to your team.

1. We ask ChatGPT the questions your customers ask

For each category we keep a fixed set of about a dozen real buyer questions, the same kind a shopper actually types. We never mention your brand or your competitors in the questions, because we want to measure whether ChatGPT brings you up on its own, not whether it can repeat a name we fed it.

We run those questions through OpenAI's ChatGPT models (with web search on, so the answers reflect real, current sources), and we save every answer.

2. We read the answers into structured data

We extract, from each answer, which brands were mentioned, in what order they were recommended, which attributes the model attached to each one, and which sources it leaned on. Your brand and competitors are then matched against that list.

3. We compute the score, we don't guess it

The score out of 100 is a deterministic mix of three things: how often you're mentioned across the questions, how high you're recommended when you are, and how many of the sources shaping the answers mention you at all. Same inputs, same score, every time. The AI writes the plain-language explanation, but it never decides the number.

4. What the score is, and isn't

It's a snapshot. AI answers shift as the web and the models change, so your score can move even when you haven't. We aggregate across many questions so a single odd answer doesn't swing it, and a check you re-run with the same inputs within a week returns the same saved result rather than a fresh, slightly different one.

It approximates, but doesn't perfectly equal, what you'd see typing into the ChatGPT app yourself. The app and the API pull from slightly different places. We'll always tell you the date your answers were sampled, and if any part of a check didn't complete, the report says so rather than filling the gap with a guess.

5. Which AI we track

ChatGPT today. Claude, Gemini, and Amazon's Rufus are coming, and they'll use the same method and the same report.