What Scorra measures
Where your brand stands when an AI assistant answers a buying question
What Scorra measures
When someone asks ChatGPT or Gemini which tool, bank or agency they should use, the answer names a handful of brands. Scorra measures whether yours is one of them.
Every day we put your tracked questions to 6 assistants — ChatGPT, Gemini, Claude, Perplexity, Google AI Mode and Google AI Overviews — read the answers, and record which brands were named, in what order, in what tone, and which sources the answer leaned on.
Why this is not rank tracking
A search engine returns a list. An assistant returns an opinion, and it composes a fresh one every time. Three things follow, and they shape the whole product:
- There is no position 1 to hold. The same question asked twice can name a different set of brands. One answer tells you nothing; a trend over weeks does.
- The answer is assembled from sources. If a comparison article ranks you fourth, the assistant repeats that. Changing what the assistant says means changing what it reads — which is why Scorra reports sources, not just scores.
- Yesterday cannot be re-measured. Nobody can go back and ask ChatGPT what it said last Tuesday. A day not measured is a hole in the series forever.
That last point is why measurement never stops when credits run out. Tracking is covered by your plan, not by your credit balance.
What you get out of it
A daily number for visibility, average position and share of voice, split by assistant and by question; the competitors that appear instead of you; the domains the models cite when they answer; and an audit of your own site that explains why it is or is not among them.

