--- title: "Citation Share of Voice: A Simple Score for Lean Content Teams" description: "Citation share of voice is your citations over every citation in a frozen prompt set. Name the denominator, keep engines split, and withhold a thin sample." url: "https://articles.flowcrews.com/citation-share-of-voice-a-simple-score-for-lean-content-teams" category: "Measurement" date_published: 2026-10-17 reading_time_minutes: 8 word_count: 1783 --- # Citation Share of Voice: A Simple Score for Lean Content Teams *A fail-closed percentage from the citation log you already keep—one engine at a time.* ![Citation Share of Voice: A Simple Score for Lean Content Teams](https://pub-07fb5e4955ba485b822d6b388be96d9a.r2.dev/2e474d51-a072-472e-ab2f-1a43a8990566/citation-share-of-voice-a-simple-score-for-lean-content-teams/hero-693d8fc3-51e8-4c5f-b580-de7da43e388e.jpg) **TL;DR:** - Citation share of voice is your citations divided by every citation in the same frozen prompt set, times 100. - It is not mention rate and not mention share of voice; the same week can be 40%, 20%, and 6%. - Keep engines separate. A blended percentage hides the engine you are losing. - A citation is a source link, not a brand name in the answer. - Withhold the percentage until the prompt set, domain list, and run count are frozen, and always name the denominator. ## What Citation Share of Voice Is, and How to Calculate It ## Citation Share vs AI Share of Voice and Citation Rate ## All Citations or Only Competitor Citations? ## What Is a Ghost Citation, and Does It Count? ## Should ChatGPT, Perplexity, Gemini, and AI Overviews Share One Score? ## What Sample Size Is Enough Before You Report the Number? ## Search Console Cannot Calculate Citation Share of Voice ## Which Log Fields Feed the Score, and When a Gap Should Change a Page ## Conclusion - Citation share of voice — Your domain’s citation events divided by every citation in one frozen prompt set, times 100, named with engine, prompt set, period, and denominator—not mention share of voice and not citation rate. - All citations, not a shortlist — The denominator is every citation in that set, including ghost citations (source links with no brand name); competitor-only lists, brand mentions, and unnamed mentions stay out of the percentage. - One engine at a time — Do not blend ChatGPT, Perplexity, Gemini, and AI Overviews; source overlap is low, pool size shifts, and an engine average hides the movement you need to act on. - Hold the number until the sample is real — Freeze the prompt set, domain list, and run count, compare only the same engine, set, and period, and withhold the score when the denominator or sample is too thin—one run is an anecdote. - Search Console cannot produce it — Console, Bing AI Performance, and analytics see only your site or sessions; the score comes from a prompt log recalculated on a fixed cadence, then published or withheld. - Act only on a cleared gap — Change a page when the same engine and frozen set show rivals cited and you absent across enough runs; ignore one-week swings, mention-only gains, and cross-engine averages. Freeze one engine’s prompt set, log every citation against the hold rule, and publish citation share of voice only when the denominator is large enough to name. ## Frequently Asked Questions ### What is citation share of voice? It is a source score: citations to your domain divided by every citation in the same frozen prompt set, times 100. Promptwatch’s example is 60 citations to your site out of 400, which is **15%**—and that is not the same as citation rate or mention share of voice. ### How is citation share different from mention share? A mention is the brand named in the answer. A citation is a source link. On GrowByData’s illustrative 600 responses, mention rate was **40%**, mention share was **20%**, and citation share was **6%**. Semrush found **61.7%** of domain appearances were ghost citations: linked, never named. ### Why can’t I average ChatGPT and Google AI Mode into one score? They do not draw from the same source pool. University of Toronto source-overlap scores for GPT and Perplexity were **0.096** in automotive and **0.088** in consumer electronics, where 1 would mean identical lists. A blended share hides the engine you are losing. ### What is a good citation share? There is no universal good score. At whole-engine scale, Promptwatch’s June 2026 read put reddit.com at **3.71%** of ChatGPT citations, Wikipedia at **0.56%**, and TechRadar at **0.40%**. A topic-window share can be **0.50%**—242 citations out of 48,589—and still be the honest number if you name the set, the engine, and the period. ### Can Google Search Console or Bing AI Performance calculate citation share? No. GrowByData notes those reports show only your site, not which competitors the same answers named. Share of voice has to come from a prompt log that records every citation in the set, not only yours. ### When should I withhold the percentage? Withhold it until the prompt set, competitor or domain list, and run count are frozen. SparkToro research cited by GrowByData found less than a **1-in-100** chance that two answers to the same prompt named the same brand list, so one run is an anecdote, not a refresh trigger. ## Sources - [https://promptwatch.com/glossary/citation-share](https://promptwatch.com/glossary/citation-share) - [https://growbydata.com/ai-share-of-voice/](https://growbydata.com/ai-share-of-voice/) - [https://www.empact.partners/empact-zone/ai-citation-concentration-statistics](https://www.empact.partners/empact-zone/ai-citation-concentration-statistics) - [https://www.semrush.com/blog/the-ghost-citations-study/](https://www.semrush.com/blog/the-ghost-citations-study/) - [https://nicklafferty.com/blog/ai-visibility-metrics-reference/](https://nicklafferty.com/blog/ai-visibility-metrics-reference/)