Now also serving healthcare, life sciences and MedTech. Meet K&C Vital →

Cited · Edition 2 · 7 May 2026

Cited · Edition 2

AI is citing newer content. A lot newer. The latest in GEO and AI Search, and what it means for PR.

What is Known & Cited? Known & Cited is a GEO consultancy and GEO-led PR partner. We measure how AI search cites and recommends brands, and help you change what it says.

First published as Cited Edition 2 on LinkedIn, 7 May 2026. Original LinkedIn post: [LINK PLACEHOLDER - Russ to fill in].

I still hate newsletters. But there isn’t anywhere I trust pulling GEO together for the people who actually need it. PR teams, marketers, businesses working out whether AI is recommending them or ignoring them entirely.

GEO tends to get lumped in with every other AI conversation. This one only covers GEO and what it means for PR and marketing.

Still no fixed schedule. Comes when there’s something worth saying. If it’s useful, stay. If it’s not, tell me.

If you’re wondering about Edition 1, it’s here. I posted it as a regular article rather than a proper newsletter edition by mistake. Cited lives in the newsletter from now on.

Russ


Three pieces of GEO research from the last three weeks. The headline finding is about timing.

AI search runs on a 90-day clock. Your content calendar probably doesn’t.

AirOps published The Complete AI Search Playbook for Marketers in March. ~15 million data points across AI answers and citations. The strongest finding is on freshness: content less than three months old is 3x more likely to be cited in AI search. Pages older than three months are 3x more likely to lose AI visibility. 70% of AI-cited pages were updated in the past year. For SaaS, finance and news, the window is “tighter still.”

Second finding worth a thought: 68% of brand mentions in AI search appear in only one model. Measure ChatGPT only and you’re seeing about a third of your visibility picture. Engines disagree about which sources to cite far more than they disagree about which brands to recommend. The cleanest read needs more than one engine, and a content plan tighter than annual.

Read the full story: https://www.airops.com/report/ai-search-playbook-marketers

Seer Interactive’s GEO Olympics is the cleanest single-dataset measurement work the field has produced.

Six LLMs, every category in the 2026 Winter Olympics, 231,347 LLM responses. Headline: a 7.8x outcome gap between brands with strong “signal architecture” (entity authority, third-party validation, community discussion) and brands without. About 1 in 5 LLM mentions of a brand are framed in stale or out-of-date language. In short: you’re either in the consideration set, or you’re invisible. Real-world performance does not carry over to AI surfacing on its own.

Read the full story: https://www.seerinteractive.com/insights/geo-olympics-study

Brinker just put GEO on the official MartechMap. The category is now real.

Scott Brinker (chiefmartec) and Frans Riemersma published State of Martech 2026 on 5 May. They formally renamed the SEO subcategory of the official MartechMap to SEO/AEO/GEO, gave it three pages, and laid out a working playbook. We’re including this not because it changes anything you do this quarter, but because it’s the moment GEO becomes a recognised category in the official record. Worth knowing. The discipline now has a name in the place names get made official.

Read the full story: https://chiefmartec.com/2026/05/state-of-martech-2026/

What we’re seeing at Known & Cited

Two things from this fortnight.

The first: the same methodology question keeps coming up. Sceptical buyers and partner agencies, a few times a week now: why does the standard AVS measure three AI engines, not five? Short version. ChatGPT, Google AI Overviews, and Perplexity cover roughly 95% of where buyers actually run AI search today. Adding Claude, Gemini, or DeepSeek lifts coverage by single digits. The standard read is the buyer read. We can scope additional engines if a client needs them. We wrote the long answer down because it kept coming up: https://knownandcited.com/blog/why-three-ai-engines-not-five

The second: the conversation mix has shifted. This fortnight, a PR agency working out whether GEO measurement belongs in their service stack, a mid-market B2B running a methodology audit before signing, a charity comms lead trying to work out whether their cause shows up in AI answers, and two consultancies asking about partner referral terms. The top question across all of them is the engine-coverage one.

If you want to see where your business stands, the AVS Exec Brief is free (and far better than the AVS Flash Report that we binned off this week: https://knownandcited.com/avs-exec-brief

Not AI

Two weeks from now I’m speaking on a panel at what is apparently the second most important health conference in the world. I had no idea the second most important health conference in the world was a thing, never mind that I’d be on stage at it. No pressure.

A few other bits have landed in the same fortnight that K&C usually does for clients rather than on its own behalf. An invite to write a piece for one of the Guardian’s supplements, and a long conversation with a really good journalist at Business Insider that might turn into something. I’m finally achieving PR results for my own thing, based on a GEO strategy... IT’S WORKING (for now at least...)

Be Known. Be Cited.

See you next time.

Read next:

Want to know where your brand stands?

The AVS Exec Brief measures whether AI is citing you, where, and how often. Free. 48 hours.

Start AVS Exec Brief