Citation share is your domain's slice of the total URL-citation pool AI engines produce across a tracked prompt set. Where share of AI voice counts brand mentions (linked or not), citation share counts only the linked citations: the URLs AI engines actually attribute their passages to. It is the competitive metric that answers "when AI cites somebody in my category, how often is that somebody me?" Track it alongside citation rate (absolute performance) and citation velocity (momentum).
What is citation share?
Citation share is a percentage. Numerator: the count of linked URL citations pointing to your domain across a tracked answer set. Denominator: the total count of linked citations across all domains in the same answer set. The ratio is your slice of the citation pie.
The metric has converged across the major AI visibility platforms. AirOps defines it as "percentage of total citations in a defined prompt set that reference your brand, vs competitors." Profound treats it as "share-of-voice style metric, portion of answer-engine citations that point to your domain vs competitors." Superlines frames it as "proportion of total AI citations a domain receives within a competitive category." AuthorityStack describes "relative percentage of AI responses naming your brand vs competitors." Same shape, slight framing differences, one canonical meaning.
The distinction from share of AI voice is narrow but material: SoAV counts every mention of your brand name, linked or unlinked. Citation share counts only the linked URL attributions. A brand mentioned in ten AI answers but never linked has strong SoAV and zero citation share. In 2026, citation share is the more honest indicator of downstream pipeline value because only linked citations drive AI referral traffic.
How citation share is calculated
The formula
Citation share = (citations to your domain / total citations across all domains) × 100. Unit: percent. The scope of "total citations" is whatever answer set you define: a single topic cluster, a category prompt set, a mixed brand-plus-category set, or a cross-platform combined view. The answer set is the universe; your share is your slice of it. A note on naming: many AI visibility tools call this metric Share of Citation, and some compute it response-based (the share of AI answers that cite you at all) rather than pool-based. That response-based figure is closer to citation rate, so confirm the denominator before comparing numbers across tools.
Per-platform vs aggregate
Citation share calculated across ChatGPT, Perplexity, Gemini, Claude, and Google AI Overviews combined gives you a portfolio view. Per-platform splits matter too. A brand at 20 percent aggregate citation share can be at 35 percent on Perplexity and 5 percent on ChatGPT, which is a different problem to solve than uniform 20 percent across all engines. Report both the aggregate and the per-platform splits.
Self-citations and counting rules
Two accounting decisions matter. First, self-citations: when an AI engine cites multiple pages from your domain within one answer, most platforms count each distinct URL once. Don't count your own internal link network as inflating citation share. Second, cross-domain counting: if a brand owns multiple domains (brand.com, docs.brand.com, community.brand.com), decide upfront whether those roll up under the same "domain" for measurement. Most tools treat subdomains as distinct; rollup logic is configurable.
Citation share vs related metrics
Citation share, citation rate, and share of AI voice all sound similar and measure different things.
The cluster works together. Citation rate tells you how often you show up at all. Citation share tells you how big a slice you capture when AI is citing somebody. Share of AI voice tells you whether the AI is even talking about you (linked or not). Velocity and decay tell you whether your position is trending up or down. One metric alone is misleading. Read them as a dashboard.
Why citation share matters
It is competitive by construction. Unlike citation rate (which can grow or shrink independently for each brand), citation share is zero-sum within a tracked set. When your share grows, someone else's shrinks. For category leaders, this makes citation share the clearer defensive metric. For challengers, it makes the displacement strategy explicit: whose citations do we want to capture and why.
It is a cleaner pipeline predictor than mention-based metrics. Only linked citations drive AI referral traffic. Only pages that are actually cited appear in the click-through paths that modern AI platforms surface. A brand with high SoAV but low citation share is winning awareness and losing pipeline. The 7-step citation-share playbook walks through the mechanical fix.
It is a shared scoreboard across engines. The formula holds on ChatGPT, Perplexity, Gemini, Claude, and Google AI Overviews with no adjustment. Per-platform citation share splits are often the fastest way to spot which engine is under-representing your brand and why, which lets optimization work focus where the gap is, not where the total pool is largest. See Answer Engine Insights for the per-platform view.
How to improve citation share
Four levers, ordered by speed-to-impact.
1. Technical access
Citation share is zero for pages AI crawlers cannot reach. Run the AI Bot Access Checker to verify GPTBot, ClaudeBot, PerplexityBot, and Google-Extended can crawl your site. Publish an llms.txt file. Confirm no Cloudflare or WAF rules are silently blocking AI user agents. This layer is binary: fixed or not. Fix it first.
2. Content extractability
Pages that match more of the query fan-out cluster earn more citations. Direct-answer paragraphs in the first 30 percent of the page capture retrieval on the head intent. Question-format H2s mirror likely sub-queries. FAQPage schema turns Q and A blocks into directly-quotable citation units. Comparison tables and definition blocks give AI engines discrete units to extract.
3. Topical depth
Single flagship posts rarely capture large citation share in competitive categories. Topic clusters (one pillar plus 5 to 12 supporting pages) cumulatively match more sub-queries in the fan-out set than any single page can. Topical authority has a 0.65 correlation with AI citation frequency per industry research: depth is a ceiling-raiser.
4. Off-site presence
AI engines cite third-party sources at high rates: Reddit, industry publications, review platforms (G2, Capterra), Wikipedia, and trade press. Citation share moves when your brand starts showing up inside those third-party sources that AI is already citing. This is the slowest lever (60 to 90 days to measurable effect), but often the highest-leverage one in crowded B2B categories.
Common misconceptions
Citation share and share of AI voice are the same
They are not. Share of AI voice counts brand mentions (linked or unlinked); citation share counts only linked URL citations. The gap between the two is often the clearest indicator of how much "known but invisible to the click path" awareness a brand has built. Brands with high SoAV and low citation share typically have a brand-to-pipeline conversion problem: AI engines describe you but do not attribute passages to you.
Higher citation share always means better pipeline
Only if the cited pages are pages you want cited. Share earned by outdated, low-quality, or off-brand pages can produce referral traffic to the wrong place. A brand at 18 percent share where most citations point to pricing, product, and high-intent comparison pages is in a different situation than a brand at 18 percent where citations cluster on stale blog posts. Segment citation share by landing-page type, not just by domain aggregate.
Citation share is stable week to week
It is not. AI retrieval is probabilistic and citation sets drift. A 2 to 4 percentage-point band of week-over-week movement is normal noise. Sustained movement outside that band across multiple weeks is signal. Same measurement discipline as citation rate: trend lines matter more than single snapshots.
Improving citation share requires more content volume
More often it requires better content. One well-structured cluster that matches 30 fan-out sub-queries outperforms ten shallow posts that match 3 each. Extractability and topical depth move citation share; volume alone does not. The exception is programmatic AEO for long-tail categories, where volume is the entire play but every page is still structured for extraction.
Frequently asked questions
#What is citation share in simple terms?
Citation share is your slice of the AI-citation pie. For a tracked prompt set, count every URL citation AI engines return across all brands in the category, then measure what percent point to your domain. If AI engines produced 400 total citations in your tracked answer set and 64 of them were yours, your citation share is 16 percent. It is a competitive metric: share of AI voice measures brand mentions, citation rate measures how often you get cited, citation share measures how much of the total citation pool you capture.
#How is citation share different from share of AI voice?
Share of AI voice counts mentions. Citation share counts linked URL citations. AI engines frequently mention brands by name without citing a URL (a paragraph describing your product but not linking to your domain). SoAV counts those. Citation share does not. The difference matters because linked citations drive AI referral traffic and attributable conversion; unlinked mentions build awareness but not pipeline. Track both, but optimize for citation share when the goal is pipeline.
#What is a good citation share benchmark?
Context-dependent. In crowded B2B categories, 5 to 15 percent citation share on non-branded category prompts is a normal competitive starting range. 20 percent and above is category leadership. In emerging or niche categories where the cited-source pool is smaller, 25 to 40 percent is achievable for brands with strong topical authority. The more useful metric is trajectory: is your share growing or shrinking against your competitor set. Single-point benchmarks are directional at best.
#Is this the same as citation share in academic research?
Related but different. In bibliometrics, citation share describes how citations are distributed across authors, papers, or journals within a field (author X accounts for Y percent of total citations in topic Z). The AI-marketing meaning borrows the mechanic (share of a total citation pool) but applies it to AI-generated answers and tracked prompt sets instead of scholarly publications. If you arrived looking for scholarly bibliometrics, tools like Clarivate Web of Science or Scopus are the resources you want.
#How do I increase my citation share?
Four levers, ordered by speed-to-impact. First, technical: make sure AI crawlers can reach your site (robots.txt, schema, llms.txt). Second, extractability: direct-answer paragraphs, FAQPage schema, question-format H2s so your pages match more fan-out sub-queries than competitors'. Third, topical depth: a pillar plus 5 to 12 interlinked supporting pages cumulatively captures more citations than single flagship posts. Fourth, off-site: get mentioned on Reddit, G2, industry publications, and the sources AI engines already cite in your category (the 7-step Citation Share Playbook walks through this).
