AI Citation Footprint: What It Is and How to Map Yours
Your footprint is the full set of queries the AI models already cite you on — the answer-layer equivalent of your organic keyword profile. Here's what it contains, how to map it, and how it differs from share of voice.
Your AI citation footprint is the complete set of queries that AI assistants — ChatGPT, Gemini and Grok — already cite or mention your domain on. It is the answer-layer equivalent of your organic keyword profile in traditional SEO: not a single score, but the full list of buyer questions where the models name you, plus the intent behind each question, which models cite you, and the competitors named alongside. You map it with reverse AI search — starting from your domain and reading the query–domain index backwards. A footprint is the baseline you grow from: every query you are not yet cited on is a gap you can target, and every query you already win is an asset to defend.
What does an AI citation footprint include?
A footprint is not a single list of URLs — it is a stack of layers on every query where you appear. Each layer answers a different question, and together they turn a flat “you’re visible” into something you can actually work. The table below shows the layers with illustrative sample rows (not live index data) so you can see the shape of a real footprint.
| Layer | What it captures | Illustrative example |
|---|---|---|
| The query | The literal buyer question a model was asked. | “best CRM for a 5-person agency” |
| Intent | Where the question sits in the funnel, so you can prioritise. | Commercial (ready-to-evaluate) |
| Model coverage | Which of ChatGPT, Gemini and Grok named you on that query. | ChatGPT + Gemini (not Grok) |
| Cited vs mentioned | Whether you were a linked source or named in prose. | Cited (linked) on ChatGPT; mentioned on Gemini |
| Competitors named | The other domains in the same answer — your substitute set. | HubSpot, Pipedrive |
| Last seen | The most recent scan where you appeared on that query. | Scanned this week (fresh) |
Stack those layers across every query you appear on and you have your footprint: a worklist, not a verdict. The full field-by-field reading guide is in how to read your reverse AI search report.
Why your footprint matters more than a score
A visibility score collapses all of that into one number, which makes it easy to report and impossible to act on. Your footprint keeps the detail, so it doubles as a worklist: the queries you already win are assets to defend, and the queries you don’t are your AI keyword gap to close. A score moves up or down with no explanation; a footprint tells you exactly which rows changed and why, which is the difference between a status report and a plan.
It is also the natural unit for tracking change over time. When a row disappears from your footprint, that is a candidate lost citation to investigate — something a single score would silently average away.
How big should my footprint be?
This is the most common question and the most misleading one. There is no universal “good” number, and we will not invent one. Footprint size scales with two things you do not control directly: how many real buyer questions exist in your category, and how broad and well-structured your content is against them. A niche compliance tool might legitimately appear on a few dozen high-intent queries; a national retailer might appear on thousands. Comparing the two counts tells you nothing.
The comparison that is meaningful is relative: run a direct competitor through the same reverse lookup and compare footprints on the same question set. The queries they are cited on and you are not are your gap; the queries you both appear on are your competitor query overlap. That difference is actionable in a way an absolute count never is.
Footprint vs share of voice
Footprint and AI share of voice are often used interchangeably, but they answer different questions. A footprint is absolute: the actual list of queries you are cited or mentioned on, regardless of who else appears. Share of voice is relative: of all the answers in a category or query set, what proportion name you versus your rivals.
The two are sequential, not competing. You map the footprint first — that is the raw evidence. Share of voice is then a calculation on top of it: your footprint as a slice of the category’s total. A footprint with no competitive context tells you what you win; share of voice tells you how much of the room you own. Most teams need the footprint before the share-of-voice number means anything, because without the underlying queries the percentage is just another score you cannot act on.
Map yours in 3 steps
Mapping a footprint is mechanical once you have a reverse lookup. Three steps:
- Run the reverse lookup on your domain. Enter your domain into the free Domain Check. It reads the query–domain index backwards and returns every query the models have cited or mentioned you on, across ChatGPT, Gemini and Grok. The mechanic is explained in reverse domain lookup for AI citations.
- Annotate each row. For every query, note its intent, which models named you, whether you were cited or merely mentioned, and the competitors in the answer. This turns a flat list into the layered footprint in the table above.
- Compare against a rival to find the gap. Run the same lookup on a competitor’s domain and subtract. The queries they win and you don’t are your AI keyword gap — the highest-leverage part of your worklist.
Where to go from here
Once your footprint is mapped, the rest of the cluster is about working it: closing gaps, defending the queries you win, and diagnosing why ChatGPT recommends your competitor and not you where a rival out-ranks you. For the broader picture of how concentrated and volatile citations are across categories, see our State of AI Citations work.