What's the difference between AI visibility, AI rank, and citation share?

AI visibility, AI rank, and citation share are three separate metrics agencies need to report on for GEO. Visibility is the percentage of prompts where a brand appears at all. Rank is the position in the answer when it does appear. Citation share is which of the brand's sources or pages the AI engine quoted. They move independently, they each tell you something different, and conflating them is the most common mistake agencies make when they start selling GEO.

AI visibility: did you show up?

Visibility is the binary question. For a tracked prompt like "best CRM for small B2B teams", visibility asks: did the brand name appear anywhere in ChatGPT's answer, in Perplexity's answer, in Gemini, in Claude, in DeepSeek? Aggregated across a set of prompts, visibility becomes a rate: "This client appeared in 18 of 50 tracked prompts across the five engines."

Visibility is the headline metric for client reporting because it is intuitive. A digital agency with 60 SMB clients can hand a CMO a single percentage and the CMO immediately understands whether it went up or down. Visibility also moves before rank does, so it is the leading indicator of GEO work paying off.

AI rank: where did you show up?

Rank is the ordinal position within an answer that does mention the brand. If ChatGPT lists five CRMs and the client is third, rank is 3. If Perplexity lists the same client first, rank is 1.

Rank is noisier than visibility for three reasons. First, engines stochastically reorder list items between runs. Second, some answers are not ordered lists at all - they are paragraphs that mention three or four brands without ranking. Third, rank only exists when visibility is positive; if the brand did not appear, rank is undefined, not zero.

For agency work, rank is most useful as a second-tier metric: "We are visible in 14 prompts, and in 9 of those we are in the top 3." That is a more honest framing than averaging ranks across all prompts.

Citation share: which of your sources got quoted?

Citation share is the metric most agencies skip and most clients secretly care about. When Perplexity answers "what is the best B2B CRM", it cites specific URLs as sources. Citation share asks: of all the sources Perplexity cited across this client's tracked prompts, what percentage were the client's own domain, versus G2, Reddit, the New York Times, or a competitor's blog?

For an e-comm-focused growth shop, citation share is the metric that tells you whether the content work is paying off. Visibility can go up because review sites started mentioning the client. Citation share only goes up when the client's own pages are being read and quoted. Example:

Client X has 26% visibility across 40 prompts. Of the 312 sources Perplexity cited in those prompts, 14 were Client X's domain. Citation share is 14/312 = 4.5%. Last month it was 2.1%.

That is the number that justifies the content retainer.

How they move independently

A 30-person B2B SaaS agency can run a campaign that lifts visibility by 40% while citation share stays flat. That means review sites and aggregators are pulling the client into more answers, but the client's own pages are not being cited. The fix is different in each case. Visibility-only lifts come from earned mentions and entity signals; citation-share lifts come from content quality and structured data on the client's own pages.

Agencies that report all three metrics separately can have a precise conversation about what is working. Agencies that mash them into a single "AI score" can't.