How brands show up in ChatGPT, Gemini, Perplexity and Google AI Overviews, and how PR earns those citations.
How brands earn citations in ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews.
LLM reputation management helps comms teams audit what ChatGPT, Claude and Perplexity say about a brand, trace errors to their sources and correct the record.
No AI engine checks state licensing before naming a sportsbook. Why that gap in a fifty-state patchwork is becoming a liability question for operators and platforms, and where it opens a lane for regional sportsbooks national brands can't easily close.
Fanatics didn't build a better sportsbook — it converted 100 million people who already buy its jerseys. Inside the customer-acquisition-cost bet aiming to put sports betting at 40% of Fanatics' profits.
Kalshi can't out-cite Polymarket yet, so it's not trying to. Inside the CFTC-first strategy betting that regulatory clarity beats a two-year citation lead once institutional money starts choosing sides.
Polymarket turned a $3B election cycle into the highest AI citation share of any prediction market. How a blockchain platform with no odds-making desk became more retrievable than FanDuel.
A 5WPR study of 1,709 AI answers finds supplement brand NOW named 26.3% of the time, but its own site cited in just 0.1% of answers with evidence.
Five sources, the IEA, EIA, IPCC, BloombergNEF and the financial press, control what AI engines say about energy and climate. The industry is absent from its own answers.
Part of the AI Pop Culture Index, Volume 23. AI engines know individual Turkish dramas but not always the word dizi, so category recognition lags the shows.
Meta AI passed 1.5 billion monthly active users and is rolling into Facebook Search. What that means for what reporters, buyers, and Meta AI itself say about a brand.
A 64-answer pilot study of ChatGPT, Gemini, Perplexity and Claude: Particle6 appeared in 37.5% of answers, Tilly Norwood in 20.3%, and ChatGPT and Claude never named her.
AI Visibility is one signal inside the broader AI Communications discipline — the strategy, measurement, and brand-level playbooks live in the parent hub.
Read the AI Communications hubAI visibility is how often, and how accurately, a brand appears in answers from ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews. A brand can have strong press and still be absent from an answer, because engines weigh sources differently from readers. Public relations and marketing teams now track AI visibility next to coverage and search. In a Muck Rack analysis from May 2026, earned media accounted for 84% of the links cited by ChatGPT, Claude and Gemini, which makes strong PR the base of strong AI visibility. EPR’s research covers 202 articles on the topic. This hub is the home for that work, under the parent AI Communications hub.
Key facts
AI visibility is a brand’s presence in the answers people read before they click anything. It has three parts: being named, being cited as a source and being described accurately. EPR measures it as Citation Share, the portion of answers in a category that cite a brand.
Being named and being cited are different. In a 5WPR study of 1,709 AI answers, supplement brand NOW was named 26.3% of the time, but its own site was cited with evidence in 0.1% of answers. Brands that publish clear first-party evidence close that gap.
Engines lean on earned media, expert voices and primary sources. EPR’s research shows credentials count: Andrew Huberman’s podcast has about one-tenth of Joe Rogan’s audience, yet ranks nearly equal on health queries. In pharma, EPR’s coverage of the 5W AI Visibility Index shows Eli Lilly and Novo Nordisk leading among the top 25 drugmakers.
Citations rotate. Only 10.6% of cited URLs persist across 28 days, according to Digital Authority Partners, with Perplexity holding 44% and Gemini 11%. That makes steady publishing and monthly tracking part of the plan. EPR explains this in Why Do 89% of AI Citations Rotate Out Within a Month?.
EPR’s Answer Density concept puts the volume of structured content a brand needs at usually 20 to 50 pieces covering the core buyer questions. See What Is Answer Density?.
| Play | Action | Why it works |
|---|---|---|
| Earn coverage | Place stories in outlets engines cite | Earned media is 84% of cited links |
| Publish first-party evidence | Put data, sources and facts on your own site | Turns a mention into a citation |
| Back founders and experts | Attach named, credentialed voices | Credentials carry weight |
| Cover the buyer questions | Build 20 to 50 structured pieces | Answer Density |
| Track monthly | Check the five engines for your key prompts | Citations rotate |
