Share of Citation Is 13. Here's What Running a Daily AI Visibility Monitor Actually Reveals.

Share of Citation (SoC) is the percentage of AI engine responses, across a defined query set, that include your brand as a cited source. Coined by Jaxon Parrott, founder of AuthorityTech, it replaces Share of Voice as the primary AI visibility metric for brand discovery in AI-mediated search. It measures what AI engines actually do with your content, not what human audiences see.
AT's SoC is 13 as of April 11, 2026.
That number comes from the AI visibility monitor running at AuthorityTech every morning: 31 queries tracked across multiple AI engines, 561 total citation slots measured, 73 of those won by AT content. Of the 31 queries, 19 returned at least one AT citation in at least one engine's response.
What Share of Voice measures, and why it's the wrong question for AI search
Share of Voice tells you how often your brand gets mentioned relative to competitors across media coverage, social, and search. It's a metric built for a world where humans read articles and form brand opinions.
G2's August 2025 buyer research found that 50% of B2B buyers now start their purchase research inside AI tools, before they hit Google or open a browser tab. By the time they reach your website, they already have a shortlist built from AI answers.
Share of Voice tells you nothing about whether you're on that shortlist. AI engines don't weight reach; they weight citation. A mention in an article that AI engines can't parse, don't trust, or won't index gets zero citation value regardless of its media reach.
Share of Citation measures the thing that actually decides the shortlist: when a buyer asks an AI engine a question in your category, does the engine cite you?
How the monitor works, and what it actually tracks
AT's monitor runs 31 pre-defined queries across AI engines each morning. The queries span five behavioral categories:
- t1: purchase-intent questions buyers actually ask ("how to get cited in Perplexity," "which publications do AI engines cite")
- t2: bridge queries connecting earned media to AI citation outcomes
- t3: category definition queries ("what is GEO," "what is AEO")
- t4: coined term definitions ("what is Machine Relations," "what is Share of Citation")
- watch: market development signals (how the GEO agency category is forming)
For each query, the monitor counts how many AI responses include an AT citation and which slot it occupies. A "slot" is one citation in one response. If an engine cites four sources for a query, that's four slots. AT either owns one of them or doesn't.
561 total slots on April 11. AT won 73. The SoC formula: (slots won / total slots) x 100 = 13%.
The parallel "query presence" metric: 19 of 31 queries returned at least one AT citation in at least one engine's response, a 61% query presence rate based on AT's internal monitoring data.
These two numbers tell different things. Query presence is the coverage metric: are we in the room at all? SoC is the dominance metric: when we're in the room, how much of the floor do we own?
What Share of Citation reveals that Share of Voice can't
| Metric | What it measures | What it misses | Useful for |
|---|---|---|---|
| Share of Voice | Brand mentions relative to competitors across media and search | Whether AI engines actually cite you; whether your sources are AI-trusted | PR reporting, competitive media benchmarking |
| Share of Citation | % of AI engine citation slots your content occupies, by query set | Raw media reach; human engagement signals | AI-era brand visibility, earned media ROI in AI search |
| Query Presence Rate | % of tracked queries where you appear in at least one AI response | Citation depth (1 of 20 slots differs sharply from 15 of 20) | Coverage gap detection, early AI visibility triage |
Citation source quality. University of Toronto research on AI citation behavior found AI engines cite earned media five times more frequently than brand-owned content, with 82-89% of all AI citations sourcing from third-party publications. SOV counts a press release pickup the same as a Forbes feature. AI engines don't. Low SoC with high SOV typically signals a source quality problem, not a coverage volume problem.
Per-engine variance. ChatGPT and Perplexity share only 11% of their cited domains, according to AT's per-platform citation audit. A strong SoC in one engine with near-zero presence in another is a concentration risk, not a win.
Coined term ownership. The t4 category in AT's monitor tracks queries for Machine Relations, Share of Citation, and Sentiment Delta: terms AT coined. Low SoC on your own coined vocabulary means the corroboration structure isn't built yet. That's a specific distribution problem.
How to track your own Share of Citation
You don't need AT's infrastructure. You need: a defined query set (15-20 questions buyers in your category actually ask AI tools), a consistent engine set (ChatGPT, Perplexity, Gemini at minimum), and a slot counting method (count the total citations in each response, note whether your brand appears, divide your slots by total slots).
Run this monthly. The GEO paper by Aggarwal et al. (Princeton and Georgia Tech, SIGKDD 2024) measured a 30-40% improvement in AI visibility from adding cited statistics to content. Answer-first formatting, named statistics with sources, and keyword-specific headings move this number most.
What a SoC of 13 means, and what it doesn't
73 citation slots out of 561 measured, on a query set tracking AT's specific category. It's a baseline on a system that compounds.
AT's content operation has been running at full output for under 90 days. A t3_category query with zero SoC means the definition is unclaimed: that's a writing task. A t4_coined query with zero SoC means the corroboration structure doesn't exist yet: that's a distribution task. The metric maps specific work to specific gaps.
Your PR agency's monthly report almost certainly doesn't tell you any of this.
Why the metric matters: Machine Relations is the frame
Every measurement framework reflects a theory of what creates value. Share of Voice reflects a theory where reach creates awareness.
Machine Relations, coined by Jaxon Parrott in 2024, runs on a different theory: AI engines recommend brands using the same signal that created editorial credibility with humans. Earned media placements in publications AI systems already trust. The buyer changed. The mechanism didn't.
Share of Citation is the measurement that fits the mechanism. It asks: when a machine is mediating your buyer's research, does it cite you?
That question has a number. Most founders don't know theirs.
AuthorityTech built the measurement framework and the content system to move it. If you want to see what your current SoC looks like across ChatGPT, Perplexity, and Gemini, the visibility audit runs the same methodology.
Frequently asked questions about Share of Citation
What is Share of Citation in AI search?
Share of Citation is the percentage of AI engine citation slots, across a defined set of buyer queries, that include your brand as a cited source. A "slot" is one citation in one AI engine response. SoC = (slots your brand occupies) / (total slots measured across all queries and engines) x 100. AuthorityTech's full methodology defines the calculation in detail.
How is Share of Citation different from Share of Voice?
Share of Voice counts brand mentions across media and search relative to competitors. Share of Citation counts citation appearances inside AI engine answers to specific buyer queries. SOV measures human audience reach. SoC measures machine mediation: whether AI engines include you when constructing answers for buyers. Per Moz's 2026 analysis, 88% of Google AI Mode citations don't appear in the organic SERP top 10. A brand can have strong SOV and near-zero SoC.
How do I calculate my own Share of Citation?
Define 15-20 queries buyers in your category actually use. Pick 3 AI engines (ChatGPT, Perplexity, Gemini minimum). For each query x engine, run the query and count all citations. Track whether your brand appears. SoC = (total times your brand is cited) / (total citation slots) x 100. Run monthly. Christianlehman.com's guide to measuring AI search visibility walks through the full workflow.
About Jaxon Parrott
Jaxon Parrott is founder of AuthorityTech and creator of Machine Relations — the discipline of using high-authority earned media to influence AI training data and LLM citations. He built the 5-layer Machine Relations stack to move brands from un-indexed to definitive AI answers.
Read his Entrepreneur profile, and follow on LinkedIn and X.
Jaxon Parrott