I had a client last year who genuinely believed their SEO had stopped working. Rankings were fine, positions hadn’t moved, and yet enquiries had dropped noticeably. What had actually happened was that Google had started showing an AI Overview above their result for their most valuable search terms, and a competitor, not them, was the business being named and described inside it.
Their ranking hadn’t broken. Their visibility inside the answer had simply gone to someone else.
I run Essheo, a search marketing agency working across the UK and US, and this scenario has become one of the most common things I help clients understand and fix. AI Overviews aren’t a future concern any more.
They’re already reshaping which businesses get seen and chosen across huge swathes of Google search, and I want to walk you through exactly how they work, what the data actually shows, and what your business needs to do about it.
What Exactly Are Google’s AI Overviews and How Big Is Their Reach?
AI Overviews are the AI-generated summaries Google displays at the top of search results, synthesising an answer from multiple web sources rather than simply listing links for a user to click through themselves. They launched broadly in 2024 and have expanded aggressively since, now running on Gemini 3 as their default underlying model globally.
As of early 2026, they were appearing on roughly 48% of all Google searches according to research into local search impact, with other analysis putting business-intent search coverage as high as 86.7%.
The reach is now genuinely enormous. Google’s own reporting puts AI Overviews at around 2 billion monthly users across more than 200 countries, a figure that grew further once Google began rolling generative UI features, including custom calculators and interactive tools, out of AI Mode and into AI Overviews from 19 August 2026, expanding potential reach to approximately 2.5 billion monthly users.
For any business relying on Google as a primary discovery channel, that means close to half of relevant searches, and a growing majority of commercial ones, now show an AI-generated answer before a single traditional result.
Do AI Overviews Actually Reduce Clicks to Websites?
Yes, measurably. Google’s AI Overviews cut organic click-through rates by up to 46% according to research cited in local search studies, and separate longitudinal analysis found organic click-through rate falling from 1.62% to 0.61% on queries where an AI Overview appears, a 61% relative decline.
The important nuance, though, is that this happens because AI Overviews genuinely satisfy a portion of searcher intent on the page itself, not because your ranking has weakened. Impressions can stay flat or even rise while actual clicks fall, since your appearance inside an AI Overview still counts toward impression data even when the user never visits your site.
There’s a meaningful upside buried in this though. Businesses cited as a source inside an AI Overview see click-through rates roughly 35% higher than businesses that appear nearby but aren’t cited. Being named inside the answer, rather than merely ranking somewhere on the page beneath it, is clearly still worth considerably more traffic than being overlooked entirely.
How Does Google Actually Choose Which Sources to Cite?
This is the part I think most business owners genuinely need explained properly, because the mechanics are more knowable, and more actionable, than most assume.
What Happens Behind the Scenes When an AI Overview Is Generated?
Google’s AI Overviews retrieve sources through a two-stage process. First, your query gets broken into anywhere from eight to sixteen sub-queries through a process called query fan-out, run in parallel to gather a wide pool of candidate content.
Second, those candidate passages are scored against relevance, source authority, freshness and structural clarity, with typically three to eight sources selected for the final synthesised answer.
Crucially, Google’s system extracts answers at the passage level, not the page level. A page can rank on page fifteen of your site and still win a citation, provided the specific passage on it directly and clearly answers the exact sub-query Google generated. This is precisely why a shorter, sharply focused passage answering one specific question often outperforms a long, comprehensive page that never states its core answer plainly.
Which Factors Actually Predict Whether You Get Cited?
Research reverse-engineering more than 200 AI Overview citations identified seven structural factors, and the strongest by far is passage clarity, with a correlation of 0.84 to citation likelihood.
A passage directly answering a query in 40 to 80 words consistently beats a 3,000 word article that buries its answer somewhere in the middle. Original data, statistics or frameworks were the second strongest predictor at 0.71 correlation, and were present in 73% of cited passages across that same sample.
Entity grounding, meaning whether your brand has a clear Wikidata entry, complete Organization schema with sameAs links, and consistent name, address and phone details across the web, correlated at 0.66, with businesses carrying strong entity signals appearing in AI Overviews at 3.2 times the rate of otherwise similar content without them.
Author E-E-A-T signals correlated at 0.60, schema completeness at 0.54, and recency at 0.45, with 62% of citations in that study going to pages updated within the prior twelve months. Existing organic ranking position also matters more than some GEO commentary suggests.
A separate Ahrefs analysis found 38% of AI Overview citations came from pages already ranking in Google’s traditional top 10, showing a genuine, if imperfect, link between classic SEO strength and AI Overview visibility.
Does Location Still Matter for Local Searches?
In a more limited way than you might expect. Research from Local Falcon studying AI Overview behaviour for local queries found effectively no correlation between a business’s distance from the searcher and its ranking position within the AI Overview itself, with a correlation coefficient of just 0.001 once a business appears at all.
Proximity does still influence whether a business gets included in the first place, particularly for highly specific or urgent service queries in markets with multiple close competitors, but it barely affects ordering once you’re in.
What this tells me, and what I explain to clients in local service sectors specifically, is that once you clear the bar for inclusion, being closest to the customer stops being your main advantage. Being the clearest, best-cited business in the answer becomes what actually determines whether you’re named first.
How Are AI Overviews Specifically Affecting Local Businesses?
I think this deserves its own section because the mechanics for local, service-based queries differ in some important ways from broad informational searches.
Where Do Local AI Overview Citations Actually Come From?
Predominantly from Google Business Profile data and established local content sources, rather than pulling primarily from a business’s own website.
Analysis of this shift found that businesses appearing in AI Overviews for local queries are, almost without exception, the same businesses already appearing in the traditional local map pack. Google isn’t building an entirely separate ranking system for its AI answers. It’s summarising and citing the businesses its existing local algorithm already trusts, just weighting those signals more heavily than before.
That’s genuinely reassuring in one sense, because it means the fundamentals of local SEO haven’t been thrown out. But it also means businesses with a weak or incomplete Google Business Profile are now doubly penalised, both in the traditional map pack and in the AI summary sitting above it.
What Should Local and Service Businesses Prioritise?
A complete, accurate Google Business Profile remains the single highest-leverage lever available, with category accuracy, complete service lists, current hours and a steady flow of new photos all directly feeding how confidently Google’s AI can describe your business.
Review velocity matters more than raw review count. A business with 40 reviews collected steadily over the past six months reads very differently to Google’s systems than one with 40 reviews all dating from two years ago, since AI systems weight recency in reviews more heavily than older ranking systems did.
Citation consistency, meaning your business name, address and phone number matching exactly across your website, Google Business Profile, and every directory you appear on, remains essential, and mobile page speed and Core Web Vitals matter arguably more now, since a slow or broken page is one Google’s systems are less confident citing regardless of how good the AI answer might otherwise be.
This is exactly the kind of technical and local foundation work we prioritise early in every client engagement at Essheo, because I’ve seen genuinely strong local businesses lose citation opportunities purely because their Business Profile or NAP consistency was neglected.
What Should Your Business Actually Do About This?
Bringing the mechanics together into a practical response, I’d focus on a specific, ordered set of priorities rather than trying to fix everything simultaneously.
How Should You Restructure Your Content?
Lead with a direct, complete answer in the first 40 to 80 words of any section you want cited, using question-format H2 and H3 headings that match how a user’s query would actually be phrased, since this lets Google’s AI efficiently locate and extract the exact passage it needs.
Include a clear definition style statement early in each section, add comparison tables for anything involving multiple options, since AI Overviews frequently construct comparison-style answers, and keep sentences reasonably short so they can be summarised without distortion.
I’d also make sure your technical foundations are genuinely solid, including clean crawl paths, fast loading performance and secure HTTPS, because Google can only cite what it can properly access and parse in the first place. None of this replaces good writing or genuine expertise. It simply removes the structural barriers that stop good content from ever being considered.
How Should You Strengthen Your Entity and Authority Signals?
Build out complete Organization schema with sameAs links connecting to your verified social profiles and any Wikidata or Wikipedia presence you can genuinely earn, since entity grounding is one of the strongest predictors of citation likelihood available.
Keep your name, address and phone details completely consistent across your website and every directory, and refresh your most important content at least annually, ideally more often for anything time-sensitive, since recency continues to influence selection meaningfully.
This is precisely the kind of structured, foundations-first work we build into every client roadmap at Essheo, because the same signals that earn AI Overview citations, genuine authority, clean technical execution, and clear extractable content, also strengthen visibility across ChatGPT, Gemini and Perplexity at the same time.
We don’t treat AI Overview optimisation as a separate project from wider GEO strategy, because the underlying disciplines overlap far more than most businesses realise.
Ready to Find Out Whether You’re Being Cited or Overlooked?
If you’ve noticed impressions holding steady while clicks or enquiries quietly decline, there’s a good chance an AI Overview is now sitting above your result, and a competitor may be the business actually getting named inside it. That’s a fixable problem, but only once you know it’s happening, and most businesses I speak with have genuinely never checked.
At Essheo, every practitioner on our team carries 8 plus years of experience competing in hard, genuinely competitive sectors against household name rivals, and we rank businesses across Google, LLMs, YouTube and social platforms because Google alone is no longer a sustainable single channel to rely on.
Our clients have generated over £45 million in combined revenue over the last two years through the search marketing systems we’ve designed and implemented, built on genuine operational results rather than theory.
We’ll start with an honest audit of exactly how AI Overviews are currently treating your business across your most valuable search terms, comparing your visibility directly against your competitors, then build a tailored roadmap to get you cited, not just ranked.
Book a strategy call with me, and let’s find out exactly what Google’s AI is telling your customers about you right now.