Share of citation: the new share of voice

Share of citation is the percentage of AI-generated answers, across a defined set of buyer questions, in which an engine names or links your brand rather than a competitor. It is the answer-engine successor to share of voice, and it is becoming the number that predicts whether a brand grows or fades in a market where the first touch is a generated answer, not a page of blue links.

#What share of voice was measuring

Share of voice started as a media metric. It is your brand's presence expressed as a percentage of all the presence in a category, whether that presence is ad spend, social mentions, or organic keyword coverage. A common social version divides your brand's mentions by the total mentions across your competitive set, so 2,500 of 10,000 mentions is a 25 percent share (Sprout Social).

The reason marketers cared was never the mentions themselves. It was the link to growth. When a brand's share of voice runs ahead of its market share, the excess tends to pull market share up over time. Nielsen credits the analysts Binet and Field with quantifying this from 171 campaigns run between 1980 and 2010, finding that a brand typically gained about 0.5 points of market share for every 10 points of excess share of voice (Nielsen). Share of voice worked as a leading indicator because it tracked the surface where buyers formed impressions: the ad breaks, the feeds, the search results they actually saw.

#Why the surface moved

The surface buyers see is changing. When Google shows an AI summary at the top of a results page, people click through far less often. A Pew Research Center study of 900 U.S. adults found that users who saw an AI summary clicked a link to a website only 8 percent of the time, compared with 15 percent for users who saw a standard results page with no summary (Pew Research Center). Clicks on links inside the summary itself were rarer still, happening in about 1 percent of the searches that produced one.

The same study found users were more likely to stop browsing entirely after a page with an AI summary, ending their session on 26 percent of those pages versus 16 percent of pages with only traditional results (Pew Research Center). And this was not a fringe experience. Around one in five Google searches in the study period produced an AI summary, and 58 percent of the people tracked ran at least one such search in a single month (Pew Research Center).

Read those numbers together and the shift is clear. A growing share of buyer questions now get answered without a visit to anyone's website. If the impression is formed inside the answer, then the thing worth measuring is your presence inside the answer.

#What share of citation measures instead

Share of citation counts, across a fixed set of questions a buyer might ask an assistant, the fraction of answers that name or link your brand. The math mirrors share of voice, but the denominator changes from mentions on the open web to citations inside generated answers.

Three inputs make the metric usable:

• A question set. Pick the real questions your buyers ask, such as "best tool for X," "X versus Y," and "how do I do Z." This is your query universe. Without a fixed set, week-to-week comparisons are meaningless.
• A citation count per brand. For each answer, record which brands the engine names or links. Your share is your brand's count divided by the total brand citations across the set.
• A cadence and an engine list. AI answers are not deterministic, so a single query tells you little. Run the same set on a schedule across the engines your buyers use, such as ChatGPT, Perplexity, Gemini, and Google's AI answers, and track the trend rather than any single response.

The output is a percentage you can watch over time and split by engine, by question cluster, and by competitor.

#Why it is a better leading indicator now

Share of voice predicted growth because it measured the surface where consideration happened. That logic has not changed. What changed is the surface. In answer engines, consideration is gated twice: first the engine has to retrieve your page, then it has to choose to cite it in the answer. A page can rank, get crawled, and still never appear in a single generated answer. Traffic and rankings miss that gap entirely. Share of citation is built to measure it.

There is also a scarcity effect. When a buyer scans ten blue links, being fourth still earns some attention. When an assistant names three sources in a paragraph, the brands outside that set are not ranked lower, they are absent. Fewer slots means each citation carries more weight, which is exactly the condition under which a presence metric becomes a sharp leading indicator rather than a soft one.

#How to start tracking it

You do not need a platform to begin. Start with a spreadsheet.

1. Write down 20 to 40 real buyer questions. Group them into clusters such as category, comparison, and how-to.
2. List your competitive set. The brands you expect to compete with for a citation, plus a bucket for "someone else."
3. Run the set across your target engines on a fixed schedule, for example the first business day of each week, and log which brands each answer cites.
4. Compute share of citation as your citations divided by total brand citations, overall and per cluster.
5. Watch the trend and the gaps. A cluster where you are absent is a content brief. A competitor pulling ahead is a signal to inspect what the engine is actually quoting from their pages.

The last step is where share of citation earns its keep. Unlike share of voice, it points at fixable causes. If you are absent from a comparison cluster, the fix is usually a clearer, better-sourced comparison page. If a competitor is cited from a single quotable definition, you can see the shape of the answer the engine wants and write a better one.

#The one-line version

Share of voice measured your presence on the surface buyers looked at, and it predicted growth for decades because of that. Buyers now look at generated answers, and those answers cite a handful of sources or none at all. Share of citation moves the same leading-indicator logic onto the new surface: not how loud you are across the web, but how often the answer names you when it matters. Track it on a fixed question set, across the engines your buyers use, and treat every gap as the next thing to write.