What is citation rate?
Citation rate is the share of AI-generated answers, across a tracked set of queries, in which your domain is cited as a source. It is the emerging visibility metric of AI search – the closest thing answer engines have to a ranking position.
How it's measured
- Define a query set – the questions your buyers actually ask, phrased the way people phrase them to assistants.
- Sample the engines – run the set against ChatGPT, Perplexity, Gemini, Copilot and Google's AI Overviews at intervals.
- Count citations – the percentage of answers in which your domain appears as a cited or linked source.
- Track movement over time – like rank tracking, the trend matters more than any single reading, because AI answers vary between runs.
There is no standard tool or methodology yet – the metric is where rank tracking was in the early 2000s. That is precisely why the teams measuring it now have an information advantage over the teams that aren't.
What moves it
The same things that make content citable in the first place: definitive statements, verifiable data, clear entity and authorship signals, and the authority that comes from earned links. See LLM optimisation for the full practice.
Common questions
Is there a standard tool for measuring citation rate?
Not yet – the metric is where rank tracking was in the early 2000s. Current practice is sampling a fixed query set across the major engines at intervals and counting citations, manually or with emerging tools.
What's a good citation rate?
There's no industry benchmark yet. The useful reading is your trend over time and your rate relative to named competitors on the same query set.
Why does citation rate fluctuate between measurements?
AI answers vary between runs even for identical prompts, so single readings are noisy – sample repeatedly and read the trend, not the data point.