December 5

GEO vs SEO: The Search Industry’s Most Pointless Civil War

0  comments

The SEO industry has never been short of a good argument.

We’ve had two decades of link-building spats, black hat versus white hat sermons, “I’ve cracked the algorithm” LinkedIn threads and the annual Core Update meltdown.

But what’s happening now is something else. We’ve discovered a new opponent – not Google, not AI, not regulators – but ourselves.

Welcome to the GEO versus SEO apocalypse: a giant, self-inflicted distraction that’s damaging the industry far more than any economic downturn.

On one side you have the “It’s just SEO” brigade, apparently so confident in their mastery of search that you genuinely wonder why they didn’t just build OpenAI in their spare time.

On the other you have the GEO evangelists, a mixture of thoughtful practitioners who have actually tested things in generative engines, plus a healthy swarm of opportunists desperately slapping “AI SEO” on whatever they were already doing and calling it a revolution.

Both sides are mostly missing the point.

Because generative engine optimisation isn’t “just SEO”. And it isn’t magic, either.

It sits much closer to brand, PR, copywriting and content than it does to title tags and internal links. And until the SEO industry can say, in plain language, what it actually does to influence AI recommendations in buying moments, this whole “GEO = SEO” line is just a comfort blanket.

What GEO Actually Is (Hint: It’s Not a Ranking Game)

Let’s start with basics.

GEO – generative engine optimisation, is about one thing: increasing AI availability.

That’s the likelihood that a generative engine (ChatGPT, Gemini, Perplexity, Grok, whatever comes next) recommends your brand in a buying situation.

Not:

  • “Do we get mentioned somewhere in a random AI answer?”

  • Not: “Did we get cited under a non-commercial query?”

The only serious question is:

When someone asks an AI what to buy, who to hire, or which solution to choose – do we get recommended more often than our competitors?

That’s GEO.

And here’s the first major difference with SEO:

A lot of the things that move the needle in GEO do absolutely nothing for your search rankings.

Which is exactly why “it’s just SEO” is such a weak argument.

Pillar 1: Copywriting – Not Content Filling

Take web copy. Not blog spam. Not 3,000 words of keyword-stuffed fluff. I mean product and service copy: your category pages, product descriptions, service pages, pricing pages, and “why us” narratives.

In SEO, you can get away with “good enough”. Google doesn’t sit there grading the quality of your prose. It doesn’t care if your copy is emotionally dead, technically generic or utterly indistinguishable from your competitors – as long as you’re ticking the right boxes and not doing anything too daft.

Generative engines are different. When you ask:

“What’s the best X for Y?” “Which agency should I use for Z?” “What are the most fashionable leather handbags under £500?”

The model goes hunting for performance attributes – the inherent fitness signals of your product or service.

Not fluffy benefits, not generic features, but the qualities that make your offering a better choice:

  • The materials you use, and why they’re stronger, more durable, safer or more sustainable.

  • The origin, sourcing, or process behind what you sell (the steak isn’t just a steak; it’s aged for 28 days, sourced from X, cooked in Y way).

  • The way your product is constructed versus cheaper alternatives.

  • Specific use cases and outcomes that imply superior performance.

Most of this has never mattered much to SEO. You could rank quite happily with generic copy and a coherent site structure.

But for GEO, this kind of detail is precisely what a large language model uses to differentiate one product from another.

When there’s no performance information or limited difference, the model doesn’t shrug and give up. It goes back out into the network: third-party sites, reviews, forums, media coverage. It does what humans do – gathers signals and compares.

That means:

  • Investment in high-quality copywriting – performance-attribute-rich, persuasive, specific – is now a core GEO lever.

  • That investment is not justified by “it helps rankings”. It’s justified because it makes you more recommendable to AI and more convincing to humans.

And this is the first uncomfortable truth for the “it’s just SEO” crowd:

GEO is heavily dependent on something most SEOs cannot credibly claim to do: proper, trained copywriting.

Not content filling. Not “SEO copy”. Copywriting that exists to drive action and articulate difference.

If your GEO “deliverable” is shuffling keywords and tweaking H2s, you’re not doing GEO. You’re still doing SEO and trying to drag it into a new world where it doesn’t quite fit.

Pillar 2: PR and Third-Party Mentions – Brand Mentions Are the New Backlinks

The second major pillar of GEO is third-party mentions.

This is where generative engines go when your site alone can’t provide enough evidence. They look to the wider ecosystem:

  • Media coverage

  • Independent reviews

  • Customer feedback

  • Directories, roundups, “best of” lists

  • Industry commentary

In PR we’ve been saying for years that brand mentions are the new backlinks. GEO finally makes that literal.

When an AI system is trying to answer, “Which provider should I recommend?” it is, in effect, doing a rough form of brand marketing analysis:

  • Who gets talked about more?

  • Who is associated with quality, reliability, good experiences?

  • Who shows up in helpful, informative content versus low-trust junk?

None of that requires an SEO. It requires PR, comms, relationship-building and brand marketing.

Traditional SEO PR (such as it is) has mostly been about links:

  • “Can you get 10 links a month?”

  • “Are they dofollow?”

  • “We don’t care about the coverage; we just want the domain authority.”

That world does not map neatly onto GEO. The PR that matters for generative engines is explicitly:

  • Human-first: would an actual person find this story / review / mention useful or convincing?

  • Brand-centric: does it reinforce performance attributes, credibility, trust?

  • Distributed: across multiple sites, formats and contexts.

In GEO terms, SEOs don’t own any of that. You might contribute a list of target publications or outreach ideas, but the heavy lifting is PR and brand, not technical search.

Pillar 3: Content Marketing – From Ranking Machine to Influence Engine

Then there’s content.

Not “content for keywords”. Not “more blogs to catch long-tail”.

Content as advertising. Content as brand-building comms.

The idea that you create content to advertise your expertise, earn trust, and shape how people think about a problem is not new. It’s just becoming more valuable in an AI world because:

  1. Content is incredibly cheap to produce compared to traditional advertising.

  2. It creates positive feelings and familiarity towards your brand.

  3. It shapes the informational substrate that generative engines learn from.

Here there is a legitimate role for the SEO: not as the lead act, but as the intelligence unit.

SEOs can:

  • Use research and search data to understand what people are actually struggling with.

  • Map problems to products (the widget example: showing when it makes more sense to buy than DIY).

  • Guide content that explains those trade-offs in a way that an AI will later synthesise into: “Yes, you can do X yourself, but in many cases it’s more efficient / safer / cheaper long term to use [this type of product/service].”

You’re not optimising for specific prompts, you will never see most of them. You’re shaping the logic and language that AI will later join together.

That is a useful contribution. But it’s not ownership of GEO. It’s a supporting role in a bigger comms and brand strategy.

The AI Citation Fantasy

Now let’s tackle the current obsession: citations.

A whole new industry has appeared around “getting cited” by AI:

  • Chunk your content “the right way”.

  • Optimise for AI overviews.

  • Build “AI visibility” dashboards.

  • Sell this as a brand-new form of marketing effectiveness.

The economic value of this, in most cases, is close to zero.

If the citation appears under:

  • A random informational query

  • With no purchase intent

  • With no meaningful traffic

  • And no behavioural data linking it to revenue

…then what exactly are you optimising for? Ego? Screenshots? Social proof in front of other marketers?

You might get the occasional click. Occasionally, that might drive a lead. But as a primary GEO and revenue strategy, “get cited” is a weak, flimsy model.

The citation is not the goal. It is the by-product of influence.

The real question is:

“Are we shaping what AI says about our category, our solution and our brand in ways that push buyers towards us?”

If the answer is yes, the citations will happen. If the answer is no, chasing them directly is just busywork you can sell on a slide.

Google, Bothism and the AI Influencer

At this point someone will sensibly say:

“Most traffic still comes from Google.”

Correct. It does. And when Google rolls AI mode as the default, most traffic will still start there.

But that doesn’t negate GEO. Because real people behave with bothism: they use Google, and ChatGPT, and Perplexity, and sometimes Grok, depending on context.

  • Need an emergency plumber? You’re not prompting an LLM, you’re typing “emergency plumber near me” into Google.

  • Need to choose a complex B2B solution? You might Google, then ask ChatGPT, then check a comparison site, then go back to someone’s site.

  • Need ideas, frameworks, “what should I consider when choosing X?” – AI is already a starting point.

We recently landed a lead from Grok with no explicit optimisation for Grok at all. It found us via our performance attributes and the signals the wider network throws off.

GEO isn’t about “dominating one engine”. It’s about recognising that AI is now one of the biggest influencers on the internet – a super-aggregator of opinion and information – and asking a simple question:

Are we actively influencing what this influencer says about us in buying situations?

That is brand work, comms work, PR work and copy work, supported by SEO data.

So What Exactly Does the SEO Do in GEO?

This is the uncomfortable, essential question:

What are you, the SEO, actually doing between 9 and 5 that increases the chance a generative engine recommends your client in a buying moment?

Not “what could, in theory, influence things”. Not “what you can relabel as GEO on your invoice”.

What are your clear deliverables?

Because so far, the meaningful GEO levers are:

  • Performance-attribute-rich copywriting

  • PR and third-party mentions

  • Reviews and reputation building

  • Content marketing that shapes category understanding

  • Brand marketing that amplifies all of the above

All of which can be done, and are often done, without an SEO department.

That doesn’t mean SEOs are useless. Far from it. SEOs understand:

  • How people search

  • How they describe problems

  • Where intent lives

  • How information structures shape discovery

That’s valuable.

But the honest conclusion is this:

GEO is not a clever new name for the same old SEO tactics. GEO is what happens when you treat AI like a mass influencer and use principled marketing and comms to win its recommendation.

If SEOs want a seat at that table, they have to earn it by clearly describing their value:

  • Using search and prompt data to inform copy and content.

  • Helping brands prioritise where to invest performance attributes.

  • Advising on topics and angles that will most effectively shape AI narratives.

  • Measuring the downstream impact where possible.

Not by insisting, endlessly, that “good GEO is just good SEO”.

The End of Semantic Hair-Splitting

The industry can keep arguing over labels if it wants:

  • “Good GEO is good SEO.”

  • “Good SEO is just good brand marketing.”

  • “Everything is just marketing.”

You can chase that circular logic all day.

Or you can ask the only question that matters:

Can you show, in a straight line, how your work increases the likelihood that an AI system recommends this brand when someone is ready to buy?

If yes, call it what you want. If no, you’re not doing GEO. You’re just rearranging the deckchairs in Search Console.

For brands serious about generative engines, the winning team will look less like a technical SEO department and more like a mini brand unit:

  • A copywriter

  • A comms strategist

  • A PR lead

  • Supported by an SEO who brings real audience and intent intelligence to the table

That’s how you grow through content in an AI world.

Not by shouting “it’s just SEO” louder than the next agency but by doing the unglamorous, classical, effectiveness-focused work of brand, comms and copy that generative engines are quietly rewarding every single day.

Andrew Holland


Originally published on LinkedIn on 5 December 2025.


Tags

LinkedIn archive


>