GEO is the new SEO. I feel much better about that.

SEO asked whether the page could be found. GEO can ask whether the organisation is worth finding on the subject at all — which is a question I’d much rather work on.

For years content people were told that SEO had to be part of the job, and I understood why. If nobody can find the article it doesn’t matter how good it is. Search demand was real evidence that people wanted an answer, and clear structure and accessible information were never enemies of good writing.

I still never felt much enthusiasm for it.

The reason, I think, is that the work almost always began with the page. Which keyword are we targeting, what’s already ranking, how long should this be, which related questions should we cover, how many variations of the same subject can we justify publishing. The results could be genuinely useful. They could also be an extremely efficient way of producing the tenth version of something the internet already knew.

GEO is now being described as the new SEO, and slightly to my surprise I feel much better about it.

It asks a better question

The shallow version of GEO looks almost exactly like the shallow version of SEO — restructure the page, add more FAQs, mention the right entities, produce material a model can extract, find the technical trick before everybody else does. I’m no more interested in that than I was the first time round, and I suspect it will be the version that sells.

The useful version starts somewhere else. It asks what the organisation should be understood for, and whether it possesses enough evidence to deserve that understanding.

That isn’t a page question. It’s a knowledge question, and the distinction matters because an organisation can optimise an existing answer perfectly and still have nothing distinctive to say. It can make its content beautifully parseable without making the underlying claim any more credible. It can publish at enormous volume while repeating evidence available on a hundred other sites.

Sometimes the page is weak. Sometimes the knowledge behind the page is weak. SEO let you avoid working out which. GEO makes that much harder to duck.

It gives older work somewhere to land

This matters to me for a reason that isn’t especially noble.

I’ve spent years developing ideas that never had an obvious organisational home — an AI maturity curve for accounting firms, an argument about capacity created by AI being quietly absorbed rather than converted into value, work on pricing, a case for Making Tax Digital as an operating-model change rather than a compliance deadline.

All of it was grounded in research, interviews and years of writing about accountants. But to a stakeholder who didn’t know that audience deeply, it could still feel like a personal enthusiasm. Why this framework? Why this argument? Why should the business care?

The problem was rarely that the work lacked substance. It was that its value depended on somebody already understanding the audience well enough to recognise the problem it solved, and most people in the conversation didn’t.

GEO provides a more legible entry point, because most of a large business now recognises the question of what AI systems and buyers should understand about them. Once that’s the question on the table, the editorial ones stop sounding abstract. What does the evidence actually support? Where are we relying on generic claims? Which customer questions can we answer better than a competitor? Where do we need original research, or an expert conversation that would fill a genuine gap?

That’s a far easier conversation to enter than asking somebody to care about an editorial framework because I happen to find it interesting.

The method travels further than the subject

The accountancy work also risked looking smaller than the thinking behind it. Pricing, MTD and practice capacity are specific to accountants and bookkeepers, and that specificity is genuinely useful — it produces sharper questions and better evidence than any general treatment would.

But the structure underneath isn’t confined to them. A small-business audience needs better answers about cash flow, productivity and whether AI can be trusted. A mid-market finance audience needs evidence about operational change and what AI does to the finance function. A US audience asks different questions in different language and needs evidence that can’t simply be imported from Britain.

What GEO makes visible is the common shape: decide what the organisation needs to be understood for, establish what it can currently prove, find where the answer is missing or generic, gather what’s needed to strengthen it, publish that knowledge in a connected and usable form, and check afterwards whether anything changed.

The subject changes. The knowledge problem doesn’t. Which makes the work portable across audiences and geographies without pretending one body of content can serve all of them.

It puts content back at the end

This is probably the part I like most.

SEO tended to make the content asset the beginning of the process. GEO, taken seriously, makes it the result of one. Before publishing anything, an organisation may need to analyse the questions it keeps being asked, examine what competitors can actually prove, talk to practitioners, commission research, revisit old interviews, or look properly at what sales and support already know.

Only then does the article or guide or knowledge page have much reason to exist.

That is closer to the journalism I was originally taught — not the mechanical part of writing the story, but the part where somebody works out what’s missing, finds the people who know, tests the claim and decides what the evidence means.

Production still matters, and so does technical accessibility. A good answer buried in an unsearchable PDF helps nobody. But the optimisation follows the knowledge rather than substituting for it.

It could easily become the thing I disliked

I shouldn’t get too comfortable, because every strategic discipline eventually acquires a checklist.

GEO could become a market for audits, scoring tools, synthetic FAQs, citation tricks and large volumes of supposedly AI-friendly pages. Organisations could respond to weak visibility by producing more material without once asking whether any of it contains knowledge worth finding. At that point GEO would have recreated the content economy SEO produced, with a different crawler in mind. Given how the last cycle went, that’s the way to bet.

There’s a second risk that’s more specifically mine. An AI maturity curve doesn’t become correct because it can be connected to GEO. A pricing argument doesn’t become evidence because a model fails to mention it. GEO can expose a gap and give existing work strategic relevance — it can’t retrospectively validate the answer, and I’d be relabelling rather than thinking if I treated it that way.

So I don’t want to rename every piece of editorial work as GEO. I want to use it to ask harder questions of the work.

The honest summary is that I feel better about GEO because, at its best, it rewards the things I’d already decided were the interesting part: evidence, organisational memory, original research, customer language, connected narratives, and finding out what an organisation doesn’t yet know.

That’s convenient enough that I should be suspicious of it. GEO may well develop in the other direction — the tactical version is easier to sell, a dashboard demonstrates better than an improved evidence base, and a visibility score moves faster than organisational knowledge ever will.

But it at least creates the opening. SEO asked whether the page could be found. GEO can ask whether the organisation is worth finding on the subject at all, and I’d much rather spend my time on the second question.