Preview Post
- May 3, 2025
Search is changing, and not quietly. Google’s AI overviews are now mostly rolled out across core search results in the USA and UK. They’ve said they have already served billions of search queries. This isn’t a Labs experiment anymore. It’s a structural change in how users interact with Google and how it plans to monetise those interactions.
I’ve worked in SEO for over a decade. I’ve seen all the noise before. This one matters.
AI overviews now appear directly in the search results, especially for complex or multi-layered queries. They pull together information, summarise it, and answer the question directly in the SERP. In most cases, the user doesn’t even need to scroll.
That has consequences. Kevin Indig’s analysis, based on Similarweb data across billions of queries, shows it clearly: Visits to Google are up, but time on the site is down. Users are in and out quickly. Page views on publishers websites have dropped after launch and only started recovering once users adapted. Query length barely moved. This is what it looks like when a platform teaches users that a one-box answer is good enough.
| Claim | Status | Why |
| “Search usage has increased” | Partly true | Visits are up, but time-on-site and depth are down. |
| “People are using longer, more complex queries” | False | Query length changes are minimal or negative. |
Image Credit: Kevin Indig
This isn’t about improving search for the sake of it. It’s about commercial alignment. Sundar Pichai confirmed in the 2025 Q1 earnings call that Google is already testing ads above and below AI overviews. It’s not theoretical. It’s happening now.
Informational queries have historically been difficult to monetise for Google. AI overviews may change that. Google gets to answer the question, hold the user’s attention, and control the space around it for ad placement. They’re creating a tighter, more contained version of the search experience that works for both users and revenue.
| Area | Key Takeaway |
| Product Strategy | AI Overviews are now part of mainstream search, focused on complex queries. |
| Cost & Infrastructure | SGE is now far cheaper to serve, thanks to TPUs and Gemini models. |
| Monetisation | Ads are being integrated into AI responses, similar to prior Search UX evolutions. |
| Traffic Concerns | Google is carefully managing rollout to maintain site traffic and address publisher concerns. |
| Engagement Metrics | Google claims increases in usage and satisfaction, but without detailed public metrics. |
If users are visiting more often but spending less time, content needs to be sharper, faster and more intentional. Being visible now means more than ranking. It means being used, cited, or referenced as part of Google’s own output.
What I’m focusing on:
Chasing rankings is the wrong mindset here. The right question is whether your content is good enough to be the foundation of Google’s own answers.
I’ve heard “SEO is dead” since 2012. It isn’t. It’s just evolving. AI overviews change how search works, but they don’t remove the need for high-quality information. Google still needs content. It still needs sources. It still needs the real web.
If your strategy is based on shortcuts or gimmicks, this shift will hurt. But if you’re working with businesses that actually know their subject and building content that reflects that, there is still plenty of opportunity.
AI overviews are now part of the default Google experience. They will reduce website clicks for some queries. They will shift how people engage with the SERP. They will open up new ad real estate for Google. But they also open up a new form of visibility for content that deserves to be surfaced. It’s not about beating the algorithm. It’s about earning your place in the output – which it is what is has alway been about.
Comments: