There’s a new acronym taking money out of SEO budgets in 2026: GEO, or generative engine optimisation — the promise that a specialist can get your brand reliably cited inside AI answers like Google’s AI Overviews, ChatGPT, or Perplexity. It sounds like the essential next frontier, and the pitch is compelling: AI is where attention is going, so pay us to make sure the AI recommends you. Here’s the uncomfortable truth the pitch leaves out. Almost nobody can do this reliably. By the industry’s own surveys, only about one in ten link builders claims a repeatable, dependable process for earning AI citations. The other ninety percent are guessing. Which means a great deal of what’s being sold as GEO expertise is selling certainty in a game nobody has actually cracked — and the sharper operators selling it know that perfectly well.

The Number That Gives It Away

Start with the statistic, because it reframes the entire market. When the link-building industry surveys itself on AI, the finding that keeps surfacing is that only around eleven percent of practitioners have a repeatable, reliable process for getting cited in AI-generated results. Read that the other way round: nearly nine in ten people working in this space cannot dependably reproduce an AI citation. This isn’t a fringe skill gap — it’s the overwhelming majority of the industry admitting the outcome is not something they control.

Now hold that against what GEO is sold as: a service that will get you cited. If nearly ninety percent of the field can’t do that on demand, then most GEO offerings are, at best, informed experimentation dressed as a reliable service, and at worst simply selling an outcome the seller cannot deliver. The confident packaging — ‘our GEO process gets you into AI answers’ — sits directly on top of an industry-wide admission that the process isn’t reliable. That gap between the pitch and the survey is the whole story.

The One Thing That Actually Correlates

If reliable AI citation is mostly unsolved, is there anything that does predict it? Yes — and it’s the detail that quietly dismantles the case for GEO as a separate discipline.

AI Cites What Already Ranks

The most consistent finding across 2026 research is that AI systems overwhelmingly cite pages that already rank in the traditional top ten — by some analyses, around three-quarters of the time. That’s the strongest, most repeatable predictor of AI citation anyone has found: not a special file, not a clever schema trick, not a GEO consultant’s secret method, but ranking well in ordinary search. Which is exactly what showing up in AI search answers comes down to — if you’re in the top results, you’re in the pool the AI draws from; if you’re not, no amount of ‘AI optimisation’ conjures you into the answer.

Which Means GEO Is Mostly Just SEO

Follow that logic and the conclusion is unavoidable. If the reliable route to being cited by AI is ranking in the top ten, and ranking in the top ten is driven by the same authority, relevance, and links it always has been, then ‘getting cited by AI’ is mostly a by-product of doing SEO well. The reliable part of GEO is just SEO wearing a newer, more expensive name. The genuinely AI-specific tactics — the part that would justify GEO as a separate paid discipline — are precisely the part nobody can do reliably. Strip out the SEO that was already working, and what’s left is the ninety-percent guesswork.

Why The Pitch Sells Anyway

If GEO is mostly repackaged SEO plus a large helping of uncertainty, why is it such an easy sell? Because it pushes exactly the right psychological buttons in an anxious market.

• Fear of being left behind. AI feels like a revolution you can’t afford to miss, so ‘AI-specific’ services sell themselves on the dread of falling behind competitors — the same fear that sells every new acronym before anyone has proven it works.
• It sounds like it needs a specialist. ‘Generative engine optimisation’ sounds technical and new enough that surely you can’t just do it yourself. The jargon manufactures the sense that expertise is required for something that, in its reliable form, is work you were already doing.
• The outcome is hard to audit. AI answers vary by user, phrasing, and moment, so when a citation doesn’t appear, it’s easy to blame the model’s unpredictability rather than the service. That un-auditability is convenient for anyone selling an outcome they can’t actually guarantee.

None of this means AI visibility doesn’t matter — it does, and it’s worth caring about whether your brand shows up when people ask an AI for a recommendation. It means you should be deeply sceptical of anyone selling a reliable shortcut to it, because the industry’s own numbers say that reliable shortcut doesn’t exist yet. The honest version of the service is ‘we’ll build the authority that ranks you, which is the best-known route to being cited too’ — which is, again, just good SEO.

What To Actually Do

If you want the AI-visibility outcome without paying a premium for guesswork, the move is to invest in the thing that actually correlates with it:

• Rank well in classic search. Since AI cites what ranks, getting into the traditional top ten is the most reliable AI-visibility strategy there is. Everything that has always driven rankings — authority, relevance, links — is the real work here.
• Build genuine authority and links. The link profile that ranks you is the same asset that makes you an AI-citation candidate. This is where budget produces measurable returns, versus GEO tactics that mostly don’t.
• Treat GEO claims with the survey in mind. If someone promises reliable AI citations, remember that roughly ninety percent of the industry admits it can’t deliver that. Ask exactly what’s AI-specific about their process — and whether it’s anything more than the SEO you could invest in directly.
• Don’t divert budget from what works. Money moved from proven link building into unproven GEO is money moved from the thing that reliably drives both rankings and citations into the thing that reliably drives neither. Keep it where the evidence is.

For most site owners and agencies, the smartest ‘AI strategy’ in 2026 is unglamorous: keep building the authority that ranks you. Solid top quality link packages and a genuine, diverse profile do more for your AI visibility than any GEO package, because they put you in the top-ten pool the AI actually pulls from. And if you manage clients, link building for SEO agencies lets you deliver that real, evidence-backed work rather than reselling an outcome nobody can guarantee.

The Bottom Line

GEO is being sold as a reliable way to get cited by AI, but the industry’s own surveys say only around one in ten practitioners has a repeatable process for it — meaning most GEO offerings promise an outcome the seller can’t dependably produce. The one thing that genuinely correlates with AI citation is ranking in the traditional top ten, which is driven by the same authority and links as ever — so the reliable part of GEO is just SEO with a costlier name, and the AI-specific part is the guesswork nobody has solved. The pitch sells because it exploits fear of missing out, technical-sounding jargon, and an outcome that’s hard to audit. Don’t fall for it. Put your budget into the authority that ranks you — buy premium backlinks for SEO that actually move the needle, or Google safe backlinks for sale and build the profile that gets you into the pool AI cites from. When someone can reliably guarantee AI citations, it’ll be front-page news. Until then, the best AI strategy is still just doing SEO well.