AI Overviews now show up on roughly 88% of healthcare searches and 82% of B2B technology searches in Google — but only around 4% of e-commerce product searches, down sharply from earlier in 2026. If you’ve been told “GEO matters for everyone equally,” that gap is the reason to be sceptical. The truth is more useful and more specific: how much AI search visibility should shape your marketing budget depends heavily on what kind of business you run and what kind of questions people ask before they buy from you.
We track this because it changes what we recommend to clients. A local physiotherapy clinic and a home-goods e-commerce store are both “small businesses,” but right now they need almost opposite content strategies to show up where their customers are actually looking.
The industry split nobody’s pricing into their marketing plan
Coverage of AI Overviews — the share of searches in a given category that trigger an AI-generated answer at the top of the results — varies enormously by industry. Based on the data we’ve been tracking through 2026:
| Industry | Approx. AI Overview coverage |
|---|---|
| Healthcare | ~88% |
| Education | ~83% |
| B2B technology | ~82% |
| Restaurants & hospitality | ~78% |
| E-commerce (product searches) | ~4%, down from ~29% a year earlier |
That e-commerce number is the one that surprises people. It’s not that Google has stopped caring about product content — it’s that AI-generated answers are far less useful for “buy this specific pair of shoes” queries than they are for “what’s the difference between physiotherapy and chiropractic care.” One is a decision an AI can meaningfully help with in a paragraph. The other usually ends the same way regardless of what an AI writes: the shopper wants to see the product, the price, and the reviews for themselves.
Why healthcare, education, and B2B see AI answers on almost everything
These categories share a pattern: the searches are mostly questions, not transactions. “Is X covered by insurance,” “what’s the difference between a diploma and a degree program,” “how does X software handle Y integration” — these are exactly the queries a language model is built to summarise well, pulling from multiple sources into one answer. If your business lives in this kind of query, an AI Overview is very likely to appear above your listing whether you like it or not, which means the only real choice you have is whether you’re one of the sources it’s built from.
Why product searches mostly don’t trigger one
E-commerce queries skew transactional and specific — brand names, model numbers, sizes, “near me,” “in stock.” There’s less for an AI to summarise and more for the shopper to simply see. Google appears to have pulled back AI Overviews from this territory over the past year specifically because early versions weren’t adding value on high-intent purchase queries; it was slowing people down rather than helping them. That pullback is worth knowing before you spend a content budget trying to get product pages cited in AI answers that mostly aren’t going to appear.
What this means if you sell a service
If you’re a clinic, an agency, a consultancy, a school, a B2B software company, or anything else where customers research before they buy, GEO is not optional marketing theatre — it’s sitting in front of the majority of your prospects’ searches already. The practical move is to build content around the specific questions your best clients ask before they sign, structured so an AI system can lift a clean, accurate answer out of it. We wrote about exactly why source pages get pulled into these answers — and why some don’t — in our breakdown of AI Overview citation patterns, which is worth reading alongside this if you’re deciding where to spend a content budget.
What this means if you sell a product
If you run an e-commerce store, don’t abandon GEO — narrow it. Your product listing pages are unlikely to get cited in an AI answer any time soon, but your buying guides, comparison pages, size and fit content, and “how to choose between X and Y” pages sit in exactly the kind of informational territory where AI Overviews still show up constantly. Put your GEO effort into the content that surrounds the purchase decision, and let your product pages do the classic SEO and conversion-rate job they’ve always done. Our GEO playbook for agentic commerce goes deeper into that split if shopping-related AI traffic is a growing line in your analytics.
Ranking #1 in Google doesn’t guarantee an AI citation anymore
Here’s the twist that catches even businesses in the high-coverage categories off guard: showing up in an AI Overview used to correlate strongly with ranking in Google’s traditional top 10. Through 2026, that correlation has weakened considerably — a large share of AI Overview citations now come from pages that aren’t in the classic top 10 at all. Video platforms, community forums, and brand-owned pages that answer a question directly and concisely are increasingly pulled in alongside, or instead of, the page that technically ranks first. Content updated within the last month also earns a disproportionate share of citations compared with older, unchanged pages.
The practical implication: if you’re in a high-coverage industry, ranking well is necessary but no longer sufficient. The page also has to be written in a way that’s easy for a language model to lift a clean answer from — a direct answer near the top, clear headings phrased as the questions people actually ask, and content that’s been touched recently enough to look current.
What about restaurants, hospitality, and local services?
Restaurants and hospitality sit at roughly 78% coverage — high, but slightly behind healthcare, education, and B2B. The queries driving that number are mostly informational too: “best time to visit,” “what to order,” “is X good for a group of eight,” “does X have vegan options.” Table reservations and directions still mostly happen through Maps and direct search, not AI answers, so the split for a restaurant is similar to e-commerce’s: put GEO effort into the research-stage content, and let your booking and menu pages handle the transactional job the way they always have.
So is GEO still worth the money in 2026?
For most of the businesses we work with — service businesses, local businesses, B2B companies — yes, clearly, and the coverage numbers above are the evidence. We laid out the full cost-benefit case, including what a realistic monthly investment looks like, in our pricing breakdown for GEO in 2026. For product-heavy e-commerce, the answer is “yes, but on a smaller and more targeted slice of your content” rather than “yes, on everything.” Either way, the decision shouldn’t be made on a gut feeling about how important AI search “seems.” It should be made by checking how often AI Overviews actually appear for the specific queries your customers use — which is different for every industry, and sometimes different for every product category within the same store.
A simple test to run on your own site this week
Take your ten most valuable search queries — the ones that actually lead to a sale or a booking, not just traffic — and run each one manually in Google. Note which ones trigger an AI Overview and which don’t. If seven or eight do, GEO deserves a real content budget this quarter. If only two or three do, you’re probably better served putting that budget into classic on-page SEO and conversion work instead, and revisiting the question again in six months. We’ve seen too many small businesses buy a GEO package because a competitor did, without ever checking whether their own customers’ searches are the kind AI Overviews touch at all — don’t be one of them, and if you want a second pair of eyes on the list, our 7 GEO myths piece covers the other assumptions worth checking before you commit a budget.