The number that displaces impressions, AVE, and share of voice.
AI Communications, defined
AI Communications is the discipline of building authority across AI answer engines — ChatGPT, Claude, Perplexity, Gemini, and Google AI Overviews — alongside earned media, digital, and influencer channels. It combines public relations, Generative Engine Optimization (GEO), and AI-visibility research to influence the answers anyone asking now begins with.
Communications industries get reorganized by metrics. Public relations got reorganized by impressions. Search marketing got reorganized by rankings. The AI Communications discipline is going to be reorganized by Citation Share — the percentage of relevant AI-generated answers in a category in which an entity is named, quoted, or linked.
This is the third canonical piece in the AI Communications series, and the deep dive on Layer 6 of the Stack. The premise is simple: a metric that did not exist a couple of years ago is now the metric that defines the discipline.
Citation Share is not a vanity number. It is not a brand-lift study. It is a comparative, measurable, repeatable score of whether a brand, founder, institution, or idea appears inside the answers that buyers, voters, donors, patients, and journalists now begin their research with.
It is the new market share.
What Citation Share measures
Relevant
Citation Share is always scored against a defined set of queries — the questions a buyer, voter, patient, donor, or journalist would actually ask. Not every possible question. A defined, category-specific set, designed in advance.
AI-engine answers
The score is measured across the five major AI answer engines — ChatGPT, Claude, Perplexity, Gemini, Google AI Overviews — and any others material in a given category. Each engine has its own behavior. The number is engine-by-engine and category-by-category.
Named, quoted, or linked
Citations come in three forms. A brand mentioned by name. A founder quoted directly. A URL or publication linked as the source. Each counts.
Why it's the new market share
Market share answered a question that mattered for a century: of every buyer in a category, what percentage chose us?
Citation Share answers the question that matters next: of every question asked in a category, what percentage of the answers include us?
The shift is structural. Buyers used to encounter a brand across many surfaces. They formed an opinion over time. Today, a growing share of buyers begin in an AI answer — a single, synthesized response that names some brands and not others. The brand named in the answer is the brand considered. The brand not named might as well not exist.
There is no "second place" inside an AI answer the way there is no second place inside the first organic Google result.
Citation Share captures the only number that matters in that environment: am I in the answer or not.
How Citation Share is measured
Step 1: Define the query set. For a beauty brand, that is roughly forty to one hundred consumer-intent prompts. For a founder, biographical and category prompts. The set is designed in advance, locked, and re-run consistently.
Step 2: Capture answers across the five engines. Each query is run in each engine. Answers captured verbatim.
Step 3: Score citation presence. For each answer, is the entity named, quoted, or linked? Each instance counts as one citation.
Step 4: Calculate share. Citation count divided by total queries. Repeat per engine.
Step 5: Compare. Citation Share is meaningless in absolute terms. It is meaningful only against a named competitive set. The number that matters is not "we have 18%" — it is "we have 18% and the category leader has 47%."
The methodology is repeatable. Auditable. It produces a number leadership can present to a board and a finance team can understand.
What Citation Share tells you
Where the brand is invisible. Most brands assume they have AI presence because they have press, social, traffic. Citation Share reveals the inverse: substantial press, social, and traffic with near-zero Citation Share, because the engines are pulling from sources the brand has not pitched, not earned, not built.
Where competitors are dominant. The brand that ranks third in market share may rank first in Citation Share — because they invested in Wikipedia, in primary research, in trade publications, in structured authority. They are winning a battle the rest of the category has not noticed.
Where to invest. The Citation Share audit names the publications, sources, and content types the engines are actually pulling from. That is the investment map. It replaces the media plan from 2019.
What it doesn't tell you
Citation Share is directional, not absolute.
The number is sensitive to query set design. A different set of forty queries produces a different number.
The number is sensitive to engine behavior. AI engines update their retrieval and synthesis weekly. The score has to be tracked over time and re-baselined when methodology changes.
The number is not a sentiment score. Citation Share counts presence. A brand cited negatively still scores. A separate sentiment analysis is necessary.
It is the most important comparative measurement available today. It is not the only measurement.
Why every comms team needs this number
Boards are going to ask. CFOs and CMOs are starting to ask whether the company's AI presence is healthy. Without Citation Share, there is no answer.
Budget conversations require it. Reallocating a comms budget from impressions-based retainers to layer-based investments requires a metric leadership trusts.
Crisis response requires it. When an AI answer goes bad, the comms team needs to know — before the CEO does.
What changes when Citation Share enters the room
Impressions stop being the headline number. They remain useful — they correlate loosely with Citation Share, but they do not predict it. Impressions report what was published. Citation Share reports what was retrieved. The two are not the same.
The media plan changes. Pitching shifts toward the publications the engines actually pull from. The retainer scope changes. The success metric changes.
The team composition changes. Measurement becomes a senior role, not a junior analyst's job.
How to start
Run a baseline. Define the query set. Score Citation Share across the five engines. Compare to the named competitive set.
Identify the gap layers. For each layer of the Stack, identify the contribution to Citation Share. Most teams find one or two layers carrying the weight and three or four contributing nothing.
Build the recurring audit. Weekly or biweekly. Same query set. Same engines. Track the number over time. Report it to leadership the way revenue is reported.
The first baseline often surprises. Brands that assumed strong AI presence find themselves at 6%. Brands that assumed weakness find themselves at 28%. The number is the start of the conversation.
The headline
Citation Share is the new market share because the question it answers — are you in the answer — is now the question that decides whether you are in the consideration set at all.
The brands that learn to move it will win the decade. The brands that don't will be replaced by brands that did.
Ronn Torossian is the founder and chairman of 5W AI Communications, the AI Communications Firm. He is the publisher of Everything-PR and the author of two best-selling marketing books, including For Immediate Release.
