AI visibility
Brand citations on ChatGPT: read them apart from Google
AI citations and Google indexing answer different questions. Here is how to read brand mentions on ChatGPT and the other AI engines RankAI observes without mixing the two.
If you open the aiVisibility page and see your brand in a ChatGPT answer, you are not looking at a Google result. Brand citations on ChatGPT and on the other AI engines RankAI observes are a separate observation from indexing: a page can be indexed by Google and never cited in an AI answer, and the reverse is also true. Before you read any number, keep the two layers apart: on one side what Google indexes, on the other what AI engines name in their answers. The aiVisibility page exists to observe the second layer, without laying it over the first.
What aiVisibility actually observes
The public pages of RankAI name ChatGPT, Perplexity, Gemini, Grok and Google AI. These are the engines the citation observation covers. Two points matter more than they might seem: Claude and Grok are measured only after you connect them, so they should not be read as present before that step; and Gemini and Google AI Overviews are not the same observation, even if the name might suggest otherwise. Treating them as one entry means reading a figure that does not match what you are looking at.
Indexing and citation: two different questions
Indexing answers the question "is this page in Google's corpus?". A citation in an AI answer answers the question "was this brand named in this answer?". The two questions have independent outcomes. A well-indexed page may appear in no observed AI answer; a brand may be cited in an answer even when its page is not the one Google shows first. If you blend the two readings, you attribute one engine's behaviour to another.
How to read citations without confusing them with Google
The method below is an operational proposal from RankAI, not a universal standard: it is meant to make the reading repeatable, not to set a rule that applies to everyone.
- Separate the sources before you look at the numbers. Open the aiVisibility page and note, for each engine named (ChatGPT, Perplexity, Gemini, Grok, Google AI), whether the citation you are reading belongs to that engine or another one. Check: every citation you report must carry the name of the engine it comes from, and no line should mix two engines.
- Treat Gemini and Google AI Overviews as two separate entries. Do not add them together and do not compare them as if they were the same observation. Check: in your notes, Gemini and Google AI Overviews appear on separate lines, each with its own outcome.
- Check the connection status before reading Claude and Grok. These two are measured only after you connect them: if that has not happened, there is no observation to read. Check: for each of the two, the connection status is explicit before any conclusion about the presence or absence of citations.
- Keep Google indexing out of the citation reading. When you move from aiVisibility to a check on Google, the question changes: you are checking indexing, not AI citation. Check: in your report, the "indexing" section and the "AI citations" section do not share the same total or the same label.
The point that often gets missed is not technical, it is about reading: people tend to treat an AI answer as a snapshot of ranking, when it is an observation on a specific engine, at a specific moment, within a declared scope. The practical consequence is that an absent citation says nothing about indexing, and a present one does not confirm it. Our operational suggestion is to build a two-column sheet: the observed engine on the left, the citation outcome on the right, with one row for each engine named. Claude and Grok stay off the sheet until the connection has been made; Gemini and Google AI Overviews stay on separate rows. That way, when the brand appears in an answer, you know exactly which observation it belongs to, and when it does not appear, you also know what you are not measuring. It is a method for reading hygiene, not a guarantee of results: RankAI does not promise rankings, and this sheet does not produce them.
A limit to keep in mind
An API response is not a person's screen: what you read through RankAI's observation does not match what a user sees in their own browser, and it should not be presented as such. The observation also covers the engines named on the public pages, with Claude and Grok only after connection, and keeps Gemini and Google AI Overviews as separate entries: anything outside that scope is not part of this reading.
For the full picture on AI citation observation, see the in-depth page.