GEO and G-SEO

How to optimise for the AI gods

AI blog illustration of a robot holding a pen

TL;DR version:

  • AI has changed search. Generative engines don’t just list links — they give direct answers. If your content isn’t part of those answers, you’re invisible.
  • GEO helps, but only so much. Adding citations, stats, and quotes can make engines more likely to feature you. But being quoted doesn’t always mean getting clicks.
  • The catch. Too much GEO risks shallow, copy-and-paste content. One study found only about half of chatbot claims are fully backed by proper sources. Trust is shaky.
  • Enter G-SEO. A newer, intent-driven approach that focuses on clarity, structure, and usefulness. Not just getting noticed, but being trusted.
  • The smart move. Don’t pick one tactic. Blend SEO, GEO, G-SEO, and trusted voices. Think portfolio, not silver bullet.

The takeaway: stop gaming engines. Start optimising for intelligence — human and artificial. Clarity is the one optimisation that never goes out of style.

Generative Engine Optimisation (GEO) is the practice of structuring content — with citations, stats, and quotable lines — so generative AI systems are more likely to include it in their answers.

Generative Search Engine Optimisation (G-SEO) is an intent-driven approach that focuses on clarity, structure, and usefulness, making content both visible and trusted.

The Monster

AI, we’re told, has eaten everything. Blogs, SEO content, carefully keyworded thought pieces — all chewed up and spat back out as one neat chatbot answer (guaranteed to sometimes be possibly accurate in places).

You can picture the off-brand Cookie Monster (with wonky-looking hands) stuffing itself with B2B copy until nothing’s left but crumbs of human content.

But the reality is stranger. Not everything’s gone. It’s just harder to see what’s working — and what isn’t — when the lights are flickering in the search engine hall.

We warned this was coming in our piece on zero-click content: the days when visibility automatically meant clicks were already fading. You could rank first and still watch traffic flatline, because Google answered the question itself. Now generative engines have doubled down, answering every question themselves.

And marketers, gawd bless us, have rushed to find the patch. That patch has a name: Jeremy Generative Engine Optimisation (GEO).

What is GEO?

Generative Engine Optimisation — GEO, for short — is the marketer’s latest attempt to keep pace with the machines.

In short: GEO is about structuring your content so generative engines can lift and repeat it — usually by adding clear citations, quotes, and stats.

The idea is simple enough: if search engines are now generative engines, spitting out conversational answers instead of blue links, then we need to tweak our content so those engines quote us.

Princeton researchers put it to the test. They fed generative models different flavours of content and tracked what the bots regurgitated. The result? Add citations, quotations, and a few juicy statistics — and your chances of being featured jumped by as much as 40%. (Zhang et al., 2023)

Engines tend to prefer clear, inline references over footnotes* — think “According to X study (2023)…” rather than burying it at the bottom. If it’s obvious to a human, it’s obvious to the machine.

It’s SEO with a new haircut. Instead of begging Google for a ranking, you’re coaxing ChatGPT, Gemini, or Perplexity into dropping your name into their neatly packaged answers. Visibility now comes not from links, but from being the snippet chosen.

Sounds great. Except, as we pointed out in our zero-click blog, visibility isn’t the same as visits. You might get quoted — but that doesn’t guarantee anyone will click through to meet you at the source.

That doesn’t make it worthless — it just shifts the metric. Instead of traffic alone, watch for brand mentions, citations, or increased recall in sales conversations.

*Spoilsports

Is GEO just SEO with a fancy new haircut and an uncanny valley spring in its step?

Do marketers love a new acronym? YBWDB* It makes us feel like we’re in on a secret. SEO was getting a bit tired, so along comes GEO: swap one letter, fresh coat of paint.

But let’s be honest: when you strip away the jargon, a lot of what GEO tells us to do isn’t new. Be clear. Be quotable. Back up your claims with evidence. That’s just… good writing.

Irene Triendl nailed it in her B2B FAQ — sometimes our industry ties itself in knots over jargon everyone pretends to understand. “Positioning,” “messaging,” “content” — not magical incantations, just basics we overcomplicate to look clever. GEO risks falling into the same trap.

Yes, GEO has its quirks. But there’s danger in mistaking these surface tactics for strategy. Slapping a percentage in a headline isn’t the same as understanding why people care in the first place.

So, is GEO genuinely new? Or is it SEO after a makeover, hoping senpAI notices?

*You bet we do, Buckaroo**

**That’s not a real one

The GEO problem

A 2023 study on verifiability in generative search engines found only 51.5% of chatbot statements were fully supported by citations, and just 74.5% of those citations were accurate

Liu, Zhang & Liang, 2023

Here’s the snag with GEO: it can feel a bit like superstition.

Marketers swap recipes like folk remedies: “Just add an eye of citation, sprinkle some powdered quotations, and throw in a wing of stats.”

Sometimes it works.

Often it doesn’t.

And no one can quite explain why.

That’s the danger the Princeton team flagged. If everyone copies the same tricks, engines fill up with repetitive, shallow content. Worse, dodgy numbers can still slip through and get quoted, just because they look like the right format.

For users, that’s a trust issue. A chatbot can’t tell if a number comes from peer-reviewed research or a marketer’s overcaffeinated blog post. When it parrots back a claim, it looks authoritative — even if it’s anything but.

A 2023 study on verifiability in generative search engines found only 51.5% of chatbot statements were fully supported by citations, and just 74.5% of those citations were accurate (Liu, Zhang & Liang, 2023). In other words: half the time, you don’t know if what you’re reading is properly sourced at all.

The Guardian called this the “chatbot optimisation game,” and it’s not hard to see why: it feels less like search and more like chance. (Bains, 2024).

So yes, GEO can boost visibility. But without care, it risks turning optimisation into manipulation — and once trust goes, everything else goes with it.

If your brand gets caught pumping out half-baked stats just to feed the machines, the hit isn’t just visibility — it’s credibility. And that’s much harder to win back.

Enter G-SEO: the grown-up version

If GEO is a quick fix, G-SEO is the sturdier upgrade.

Earlier this year, researchers coined Role-Augmented, Intent-Driven Generative Search Engine Optimisation (RAID-GSEO) — mercifully shortened to G-SEO (Chen et al., 2025).

Instead of throwing in stats and hoping for the best, G-SEO focuses on intent: What does the user actually want? What’s the clearest way to answer it?

For example: someone searching “What is G-SEO?” wants a simple definition. Someone asking “How do I do G-SEO?” needs a step-by-step guide. Same topic, different intent, different content.

The framework breaks this down into steps: write in modular, easy-to-lift chunks; back claims with evidence; and check whether the content is genuinely useful.

In other words: GEO tries to get you noticed. G-SEO aims to make you useful and trusted. And in a world where trust is in short supply, that’s the smarter long-term play.

And no, this doesn’t mean rewriting your whole site overnight. Smaller teams can start by reworking a handful of key pages or FAQs in this style. You’ll see most of the benefit without rewriting your entire website in one go.

What about influencers?

As Lars Lofgren eloquently puts it “Influencer nodes are all that’s left after AI gobbles up everything else. When we’re drowning in an endless sea of lukewarm AI slop that was trained on a brand book (now called a Brand AI Agent) that a team put way too much time into, what will we be left with? Real stories from real fucking people.” (Lofgren, 2025)

People trust people more than faceless brand copy, and AI systems often surface content from recognisable voices. A strong personal brand still cuts through.

But that doesn’t make influencers invincible. If their posts aren’t well-structured, backed up with sources, or easy for AI to read, they risk being drowned out by slicker, more GEO-friendly content.

So maybe it’s better to think of influencers not as the “last ones standing,” but as connectors. They can shine a light on credible content for both AI engines and human audiences — but only if their insights are packaged in a way machines can understand.

Hybrid strategy

So what’s the smart move? Forget climbing a single SEO ladder. Think more like weaving a web that catches attention from different angles. (I guess attention is flies in this metaphor?)

Here are a few anchors for that web:

Keep it clear. Break content into simple, digestible chunks. Machines (and people) both prefer structure they can scan. Lists, FAQs, and tables often perform well because engines can lift them neatly — but the principle is the same: structure beats ramble.

Look past quick wins. Citations and stats help, but they’re not enough. Shape your content around what the reader is really trying to do or find.

Use influencers as connectors. They’re not the last survivors of the apocalypse. They can amplify content and build trust — but only if their insights are packaged in a way AI can understand.

Don’t ditch SEO. Old-school search still matters. Ranking in Google and showing up in ChatGPT answers isn’t overkill, it’s insurance. It’s now one part of a wider web.

The bigger point? Don’t put all your eggs in one algorithm’s basket. Spread your bets across SEO, GEO, G-SEO, and influencer voices. That way, when the rules change again — and they will — you’re not starting from scratch.

Optimise for Intelligence, not engines

The temptation is always to chase the algorithm. We did it with SEO, now we’re doing it with GEO. Add the right keywords, drop in the right citations, and wait for the machines to nod approvingly.

But if we’ve learned anything from zero-click content — and from GEO’s quick-fix recipes — it’s that chasing visibility for its own sake is a dead end. You don’t win by gaming engines. You win by making your content intelligible: to readers, to customers, and yes, to machines.

That’s why G-SEO feels like a step forward. It’s not about gaming engines, it’s about making content that actually helps. Clear answers. Structured ideas. Proof people can rely on.

Because the acronyms will keep changing — SEO, GEO, G-SEO, and whatever comes next. The one thing that doesn’t change? Clarity travels.

Clarity means short sentences, direct answers, and evidence that stands on its own.

Whether you’re writing for people or machines, that’s the optimisation that lasts. It’s a safety net.

7 G-SEO Tips (that won’t cost you your sanity)

  1. Start with intent. Don’t guess what the bot wants — figure out what the person behind the question is actually asking.
  2. Make it snackable. Break ideas into neat little chunks. If a machine can lift it cleanly, a reader can too.
  3. Prove it. Stats, quotes, sources — but real ones, not the 83% of stats you make up to reinforce your point.
  4. Say it straight. The plainer the words, the further they travel. Waffle is invisible (apologies to Belgium).
  5. Ask yourself: “So what?” If your content doesn’t actually help anyone, why would a machine repeat it?
  6. Mix your plays. SEO, GEO, G-SEO, influencers — it’s a portfolio, not a single bet.
  7. Optimise for clarity. Algorithms will change again. Clarity won’t.

FAQ: GEO and G-SEO

Q: What is GEO in plain English?

A: Generative Engine Optimisation is about writing content in a way that generative AI systems (like ChatGPT, Gemini, or Perplexity) are more likely to quote. It usually means adding clear citations, quotes, and stats that engines can lift easily.

Q: How is G-SEO different from GEO?

A: GEO chases visibility. G-SEO chases usefulness. GEO tries to get you mentioned; G-SEO makes sure your content is structured, trustworthy, and intent-driven so it’s worth repeating.

Q: Does this replace SEO?

A: No. Traditional SEO still matters for Google and Bing. Think of GEO and G-SEO as new layers on top, not a replacement.

Q: Where should I start?

A: Start small. Optimise your FAQs, help pages, or product explainers first — the content that maps most directly to user questions.

Q: How do I measure success if clicks don’t always follow?

A: Look for brand mentions in AI outputs, citations, and recall in customer conversations. Visibility has shifted — it’s not just traffic anymore.

Q: What’s the biggest risk with GEO?

A: Chasing quick wins with dodgy stats. If AI repeats unreliable claims under your name, trust in your brand takes the hit.

Further Reading

Academic & Research Sources

Industry Commentary & Explainers

Useful resources & frameworks

  • Answer Engine Optimisation (AEO): A developing framework for optimising content to appear in AI-driven answers. See Munn (2024) above.
  • Artificial Intelligence Optimisation (AIO): An emerging concept focused on improving content retrievability and trust for AI systems. Wikipedia overview
  • E-E-A-T Guidelines (Google): Experience, Expertise, Authoritativeness, and Trustworthiness remain core for both SEO and AI retrievability. Google Search Central
  • JKC Blog: Zero-Click Content (2024). A precursor to this piece, exploring how visibility no longer guarantees traffic. https://jkc.co.uk/blog/zero-click-content/

References

For those who like to check the receipts (or just enjoy a good academic rabbit hole), here are the sources behind this piece:

Write a comment


Top