Citation Share
Citation share is the percentage of AI-generated answers about your topic that name or quote your site as a source. It is the GEO equivalent of share of voice, and it is becoming the key metric for measuring how visible you actually are inside AI engines.
In old-school SEO you tracked rankings and share of voice. In GEO you track citation share. It answers a blunt question: when AI engines write about your topic, how often do they name you as a source? If the answer is 'almost never,' you have a visibility problem no ranking report will reveal.
Citation share is simply your portion of all the source citations across a set of AI answers for your target topics. Ask ChatGPT, Gemini, and Perplexity a basket of questions in your space, count how often each one cites your site versus competitors, and you have a number you can actually move and report on.
Why citation share beats raw rankings now
Rankings tell you where you sit in a list of links. But when a user gets an AI answer, they may never see that list. What they see is which sources the model chose to cite. You can rank number one and still get cited zero percent of the time, or rank number five and be the model's favorite source. Citation share measures the thing that now drives a growing slice of attention and clicks.
You can win the ranking and lose the answer. Citation share is the only metric that tells you which one happened.
How to measure it
- Build a fixed list of 20 to 50 questions that matter in your niche, phrased the way real users ask.
- Run each question through the AI engines you care about and record every source they cite.
- Count your citations against the total, and against named competitors, to get your share.
- Repeat on a regular schedule so you can watch the trend, not just a single snapshot.
- Segment by engine, because your share in Perplexity can look very different from your share in Google AI Overviews.
| Metric | What it tells you |
|---|---|
| Citation count | Raw number of times you were cited across the question set |
| Citation share | Your citations as a percent of all citations in that set |
| Share by engine | Where you are strong or weak across ChatGPT, Gemini, Perplexity |
| Trend over time | Whether your GEO work is actually moving the needle |
Example
You test 30 questions about 'small business bookkeeping' across three AI engines. Your site is cited in 9 of those answers, while the full set contains 90 total citations. Your citation share is 10 percent. Track it monthly, ship GEO improvements, and watch whether that 10 climbs. If a competitor sits at 35 percent, you now know exactly how far behind you are and have a number to close.
lightbulbPRO TIP
Keep your question list fixed between measurements. If you change the questions every time, you are measuring noise, not progress. A stable question set is what turns citation share into a trend you can trust.
targetCitation share is not vanity
Unlike a lot of soft AI metrics, citation share maps to something real: presence in the answers users actually read. A rising share means more of the AI-driven attention in your market is flowing to you. That is the closest thing GEO has to a north-star number.
Measure presence in answers
Citation share counts how often AI engines name you as a source. Track it on a fixed question set over time to prove whether your GEO efforts are working.
For the full method, including tooling and reporting, read my guide on measuring LLM citations.
Want this handled by someone who has measured search for 20 years?
Work with me