Most conversations about AEO, including several of our own posts, default to talking about ChatGPT, Gemini, Claude and Perplexity as if those four are the whole picture. They're not. Google AI Overviews, the generated summary box that now appears above traditional results for a large share of Google searches, is arguably the highest volume AI recommendation surface a small business will ever appear in, simply because it sits inside the search engine almost everyone still opens first. It gets far less dedicated advice than the standalone chat assistants, mostly because it's easy to mistake for "just Google" rather than treating it as its own distinct visibility problem.
How it's actually different from a chat assistant
ChatGPT, Claude, and to a large extent Gemini and Perplexity when used as standalone apps, generate an answer from a blend of training knowledge and, where enabled, live web retrieval triggered by the specific conversation. Google AI Overviews works differently: it's generated directly from Google's existing search index in response to a specific query, then summarised into a short answer with source links attached. That means Overviews visibility is tightly coupled to whether your page would already rank reasonably well in traditional search for that query, in a way that standalone chat assistant visibility often isn't.
Why this matters for what you actually do about it
Because Overviews draws from the existing index, the technical fundamentals that already matter for SEO, crawlability, page speed, clean HTML, valid schema, matter just as much here, and arguably more, because a page that isn't well indexed in the first place has no chance of being summarised. But indexing alone isn't enough. Overviews tends to favour content that answers a question directly and early, in a clear structure a summarisation model can lift cleanly, over content that buries the direct answer under several paragraphs of scene setting.
Where local and golf businesses tend to lose visibility here
- Content that answers questions indirectly. A page that requires three paragraphs of context before stating opening hours, pricing, or availability gives a summarisation model very little to confidently lift into a short overview.
- Thin or duplicate location pages. Multi site businesses with near identical pages per location often confuse which page should be cited for a given local query, weakening all of them.
- Missing or broken schema markup. Overviews leans on structured data to confirm facts like hours, pricing and service area with confidence. Broken or absent schema removes a signal that's genuinely load bearing here.
- Slow or poorly rendered pages. If Google's crawler struggles to render key content, that content effectively doesn't exist for indexing purposes, and therefore can't appear in an Overview either.
What to actually do about it
Treat your highest intent pages, services, pricing, location, FAQ, as if a summarisation model will lift the first two sentences verbatim, because it often does. Lead with the direct answer, then add supporting detail afterwards rather than before. Keep schema markup accurate and current, since it's one of the few signals doing double duty across both traditional search and Overviews. And test it directly: search your own key local queries in Google and read whether an Overview appears, and whether you're cited in it, rather than assuming your existing SEO position automatically carries across.
Where to start
Search Google for the exact questions your customers ask, "best [category] near [town]," "does [service] cost more at the weekend," and note whether an AI Overview appears and who it cites. If a competitor is cited and you're not, that's usually a content structure and schema gap rather than a fundamental authority gap, and it's diagnosable the same way as any other AEO issue.