What is citation rate?
Citation rate is the percentage of target questions in which an AI assistant cites the brand in its answer, across ChatGPT, Perplexity, Gemini and Claude. It is the core AI Visibility metric. You measure it on a fixed schedule, platform by platform, recording where each citation lands and how it frames the brand — as the source of the analysis, or as a passing mention.
What goes into the measurement?
| DIMENSION | WHAT IT RECORDS | EXAMPLE |
|---|---|---|
| Presence | Does the brand appear in the answer? | Cited in 12 of 50 questions = 24% |
| Position | Rank among the sources cited | 1st source vs 4th source |
| Framing | How the brand is described | "reference analysis" vs a generic mention |
| Platform | Which AI models produce the citation | Perplexity 30% · ChatGPT 18% · Gemini 10% |
| Competition | Who gets cited when the brand does not | Competitor X cited in 60% of the gaps |
Source: ReBo monitoring protocol, 2026. The figures in the example column are illustrative.
Related entries
Citation rate is the outcome GEO chases, and the yardstick that places an asset manager on the levels of the AI Visibility framework. The ReBo method measures it in step 4 and feeds the result into the next month's content roadmap.
Frequently asked questions
How is the citation rate calculated?
Define a fixed set of target questions — say, 50 questions clients would actually ask — run each one through every AI model you monitor, then divide the answers that cite the brand by the total number of questions. Track the result month by month, platform by platform.
Why does the rate move between runs with the same questions?
AI answers are not deterministic: the same question can produce different sets of sources. That is why the protocol repeats each question several times and reports the average. What matters is the trend, not any single run.
What counts as a good citation rate?
It depends on the niche and the competition. The first milestone that matters is moving off zero — existing in the answers at all. From there, the goal is steady growth on the questions with the highest commercial value, and beating direct competitors on those.