Words: Alex Coe – Digital Director
Category(s): Campaigns, Content
Date: 10th February 2026
One year on from GEO’s cat-among-the-pigeons arrival, many B2B marketers remain at an impasse. The wearisome SEO vs GEO debate continues distracting from the more useful and interesting discussion in 2026: how can the principles of SEO drive an effective GEO search strategy today?
It seems, on the face of it, a tricky one to unpick. On one hand, GEO doesn’t really exist. Google themselves have gone on record multiple times to say as much, actively advising against treating it as a separate discipline because it may harm your organic search performance. (Although Google would say that, considering it’s competitors like OpenAI who have been driving the proliferation of LLMs.)
On the other hand, from a business point of view, the market has decided that GEO does exist. Procurement teams are asking about it, competitors are claiming they do it, and agencies are expected to have a view (or potentially forfeit the work to an agency that does).
The dilemma is if agencies embrace GEO uncritically, they risk endorsing something they might not fully believe in, which can create unmanageable expectations. But refuse to acknowledge it at all and they risk sounding out of touch or dismissive of things that do feel real to customers. Plus, if GEO does become a true discipline in the future, they’ll be positioned as “anti” the whole idea, which is rarely a comfortable place to be when the market moves.
The most honest position today sits somewhere in the middle. GEO is not currently a separate discipline, but the market forces that created the conversation around it are real. Which means the sensible response isn’t to sell GEO at face value, or to pretend nothing is happening. It’s to be clear about what does and doesn’t change, and to help stakeholders understand the difference.
What GEO has changed in search strategy
On the many occasions SEO has been proclaimed to have ‘died’, the surviving practice is very obviously just SEO in a new context. Mobile, video, voice search – all had the same SEO fundamentals with a different interface. The threat from GEO is on the face of it harder to dismiss because people are genuinely using LLMs such as ChatGPT or Gemini to find information in a substantively different way.
The intent behind an LLM prompt isn’t quite the same as the intent behind a Google search. When someone searches Google, they’re often in an exploratory or evaluative mode, ready to review sources, compare options, and decide where to click next, which is why ranking in SERPs has clear commercial value.
When someone asks ChatGPT or Perplexity a question, they will often be looking for either a synthesised response they can act on, or a way to work through a problem with less effort. In this context, being referenced by an LLM positions a brand as an authoritative source or significant player rather than one option among many.
So one might argue that search engine and LLM user behaviour is substantively different, and the resulting commercial value received by a brand is different. These factors, among others, may influence a brand’s content and search strategy. But the fact remains that the fundamentals remain the same for SEO and GEO: understanding intent, creating genuinely useful content, and presenting information clearly still underpin visibility in both environments.
There’s also been a lot of noise about concepts like “chunking” content, optimising for “vectors” and “entities”, or engineering for “citations” and “brand mentions”. The GEO language feels new, but in simple terms:
- Chunking is about clear information hierarchy, which you’d do anyway for readability and structured data.
- Entities are what we used to call topical authority and semantic relevance.
- Citations are backlinks by another name, and brand mentions have been a trust signal since long before ChatGPT existed.
It’s important to keep in perspective that organic search still accounts for the overwhelming majority of discovery. So beware of emerging GEO recommendations that might actively harm SEO performance if taken too literally. Focus on strengthening the fundamentals that serve both.
Where GEO can go wrong
When someone on your team says, “we need to optimise for AI search”, they’re not wrong. What’s wrong is the idea that it requires a fundamentally new discipline.
Generative AI tools don’t discover information independently, they retrieve it. And how do they retrieve it? Well, they use search engines (ChatGPT alone accounts for 6-10% of Google’s daily search volume) then they summarise what they find and present it as though they thought of it themselves. You are therefore not optimising for an LLM directly, you are optimising for the systems that feed it, which means, functionally, you are still doing SEO.
The trouble starts when marketing teams optimise for AI rather than for the audience they serve. This is where we see things like content “chunked” into unnatural fragments that no human would ever read, which Google do not want you to do, leading to pages written to be extracted rather than experienced, viewpoints flattened because anything with a personality might confuse a model.
What happens then is that GEO-optimised content becomes actively worse for humans. And over time, that tends to backfire for the machines too, because the signals that make content useful to people are the same signals that make it worth surfacing in the first place.
Content written to be genuinely useful to your audience has always been the safest bet, and that hasn’t changed just because the interface has.
In practice, optimising for “AI search” means doubling down on the same fundamentals that have always driven sustainable visibility.
So, should you “do GEO”?
If by GEO you mean helping your brand stay visible as search interfaces evolve, understanding how AI systems source and surface information, and making sure your content is structured, authoritative, and genuinely useful enough to be worth citing, then yes. You should absolutely do that.
And if you’re trying to work out how visible your brand really is, and which problems are worth fixing, we’re always happy to have that conversation, without jargon getting in the way.