Sean Walsh
Sean is a Director at Intelligency heading up our digital marketing and client services operations. Sean has 15+ years experiencing working both in-house and agency with brands including Lloyds, Alstom, Hitachi, Lufthansa, Viaplay, DFDS Seaways and Mercedes-Benz.
There is a point at which a number stops being a data point and starts being a structural fact about how the world works. Google announced at I/O in May 2026 that AI Overviews now reaches 2.5 billion monthly users. AI Mode, its full conversational search interface launched just twelve months ago, has already crossed 1 billion monthly users. Together those two surfaces mean AI-generated content is now the default search experience for the majority of the world’s internet users.
This is the moment where the phrase ’emerging channel’ stops applying. Investing in AI search visibility is no longer a forward-looking decision. It is a response to a market structure that already exists at scale.
What the numbers actually say
Google has roughly 4.5 billion total users globally. AI Overviews reaching 2.5 billion of them means that more than half of everyone who opens Google now encounters AI-generated content before they see a traditional list of links. That is not a feature used by early adopters. It is the default experience for most of the platform.
AI Mode’s growth trajectory is if anything more significant. One billion monthly users in twelve months makes it one of the fastest-scaling search surfaces in Google’s history. Queries are reportedly more than doubling every quarter since launch. Sundar Pichai’s framing at I/O was direct: queries reached an all-time high last quarter, driven by AI features rather than despite them.
The two surfaces are distinct in important ways. AI Overviews are the AI-generated summaries that appear above traditional results on the standard search page. AI Mode is a separate, full conversational search interface powered by Gemini. Research from Ahrefs found that they cite the same URLs only 13.7% of the time. Being visible in one does not guarantee visibility in the other, and the signals that earn citation in each environment are not identical.
Outside Google, the broader landscape reinforces the same direction. ChatGPT has passed 800 million weekly active users. Around 37% of consumers now begin their searches with an AI tool rather than a traditional search engine. These numbers sit alongside, not instead of, Google’s figures. The total volume of AI-mediated search behaviour is larger than any single platform’s numbers suggest.
The click problem – and why it is less simple than it looks
The number that tends to dominate conversations about AI search is the click-through rate. Semrush data suggests that 92 to 94% of AI Mode sessions end without a click to an external website. That is a real figure and it matters. But taken in isolation it tells an incomplete story.
The more important finding, which we examined in detail when Similarweb published their downstream traffic research, is what happens after someone sees your brand in an AI-generated answer. Brands appearing in AI recommendations were 2.5 times more likely to receive a site visit within seven days than comparable brands that were not mentioned. Of that downstream traffic, more than half arrived via branded search rather than direct AI referral click. The user saw the brand in an AI answer, remembered it, and searched for it later. None of that appears in standard referral analytics.
The click-through rate on AI-generated answers is low. The downstream commercial effect of AI mentions is considerably larger than that rate implies. These are different things, and conflating them leads to under-investment in a channel that is already operating at scale.
What scale changes about the decision
When a channel reaches a billion users, the question of whether to engage with it shifts. It is no longer a question of whether the audience is there. The audience is there. The question becomes what position you hold in that channel relative to your competitors.
AI-generated answers about your sector are being generated right now, regardless of whether you have taken any steps to influence them. The only variable is whether your brand appears in those answers or whether a competitor does. At 2.5 billion monthly users, the opportunity cost of not being cited is measurable in the same terms as the opportunity cost of not ranking on page one of Google in 2010.
The concentration dynamic makes this more acute over time. Early data from AI search visibility research suggests that in established categories, a small number of brands capture the majority of AI citations, similar to how a small number of websites dominate traditional search results pages. Finance and industrial sectors remain more open, but healthcare, dental, beauty and consumer services are categories where brand concentration in AI answers is already developing. Getting into the answer set before the category settles is considerably easier than displacing an incumbent once it has.
What this means in practice
The scale of AI search adoption does not require a new strategy from scratch. The content and credibility work that earns AI citations is largely the same work that earns traditional search rankings, trust signals and brand authority. What changes is the priority weighting.
- Specificity over volume. AI systems compress multiple similar articles into a single answer. A page that answers one question clearly and completely is more likely to be cited than ten pages that cover the same topic at shallow depth. The content investment that works in AI search is fewer, better pages rather than more frequent publication.
- Authoritative third-party presence. AI systems weight third-party sources heavily when assembling answers. Being mentioned in trade publications, review platforms, professional directories and sector-relevant forums contributes to AI citation rates. Only 12% of URLs cited by major AI platforms also rank in Google’s top ten – meaning AI citation and traditional ranking are related but not the same, and the citation pathway draws more heavily on off-site authority.
- Entity clarity. AI systems need to understand what your brand is and what category it belongs to before they will cite it confidently. Structured data, consistent information across directories, a clear Wikipedia or Wikidata presence where relevant, and coherent messaging across your own site and third-party profiles all contribute to entity recognition. This is infrastructure work, not campaign work, and it compounds over time.
- Measuring the right things. Standard referral analytics will undercount the commercial effect of AI visibility. Branded search volume trends, direct traffic growth and brand mention monitoring across AI platforms give a more complete picture than referral click counts alone. Google Search Console’s recently launched platform properties feature is the first standardised tool that brings some of this visibility into the reporting stack most teams already use.
The point of no return
There is a useful analogy from the early years of mobile. For several years, having a mobile-optimised website was described as important and forward-thinking. Then smartphone penetration crossed a threshold, and it became table stakes. The question stopped being whether you needed a mobile site and became how quickly you could fix yours.
AI search is at or past the equivalent threshold. 2.5 billion monthly users on AI Overviews alone is not a pilot. It is not an experiment. It is the product. Businesses that treat AI visibility as something to consider later are not being cautious. They are ceding ground in a channel that is already being used by more people than almost any other discovery surface on the internet.
The good news is that the channel is not yet fully consolidated in most sectors. The brands that are easy to find, cite and trust across multiple public surfaces still have a meaningful advantage over those that are not. That window narrows as adoption continues to accelerate, but it has not closed. The number worth focusing on is not 2.5 billion. It is how many of those 2.5 billion users are asking questions your brand should be answering.
