Ask five market research firms how large the Generative Engine Optimization market is, and you’ll get five different answers — and they don’t agree within a factor of two.
One puts the global GEO services market at roughly $1 billion today, heading toward $17 billion by 2034. Another puts current size closer to $850 million, on a path to $19.8 billion over the same period. A third, more measured assessment puts it at $2-5 billion by 2028 and describes the market plainly: early, fragmented, and full of opportunity — comparable to where SEO itself stood in 2005.
We’ve looked at the underlying research. Nobody has settled this, because nobody can. The category is too new, the players too varied, the definitions too inconsistent for any single number to be trustworthy yet.
For a gallery owner deciding whether any of this matters, the disagreement is more useful than a confident headline figure would be. It tells you the market hasn’t been carved up. There is no incumbent. Whatever this becomes, it is being decided now — by whoever moves first, not by whoever was already biggest.
What everyone does agree on
While the size of the GEO market is contested, the underlying behavioural shift driving it is not. The data here is consistent across multiple independent sources, and it is significant.
Less than one third of Google searches now result in a click to an external website, according to clickstream research covering the first months of 2026. When an AI Overview appears in the results, that figure drops further — roughly 83% of those searches end without a click at all. Inside Google’s newer AI Mode, the figure climbs again, with the large majority of interactions resolving entirely within the AI conversation itself.
Google’s own announcement at its I/O 2026 event confirmed AI Mode had surpassed one billion monthly users, with query volume more than doubling every quarter. Worth noting honestly: AI Mode’s share of total search referral traffic remains small for now — a fraction of what traditional Google search still sends to websites. The most dramatic part of this shift has not fully arrived. But the direction of travel is not in question, and the rate of change is accelerating, not slowing.
What this means for galleries specifically
None of the market research above was written with galleries in mind. Translated into what it actually means for a gallery:
A collector researching where to buy a painting by a specific artist is increasingly likely to get their answer directly from an AI system, without ever reaching a gallery’s website. The traditional model — be found via search, win the click, convert the visitor — is being quietly bypassed for a growing share of buying journeys. The gallery that AI recommends gets the enquiry. The gallery that doesn’t appear in the answer may never know the enquiry existed at all.
This is precisely the dynamic Artist Authority is built to address: not chasing clicks in a world that increasingly doesn’t click, but building the entity authority that gets a gallery named directly in the answer itself.
Why the uncertainty favours moving now
A market nobody can size accurately is also a market nobody has won yet. The galleries that treat this as a curiosity to revisit once the picture is clearer will find, by the time it is clearer, that the picture has already been drawn — by whoever was paying attention while everyone else waited for certainty that was never going to arrive on a convenient timeline.
The data doesn’t need to agree on a number for the direction to be obvious. It already is.
robberbrand.com — building Artist Authority
