What are Google AI Overviews?

Google AI Overviews represents the search giant’s most significant interface redesign since the introduction of featured snippets. Rather than presenting users with ten blue links and expecting them to click through multiple results, Google now generates AI-powered summaries that appear at the top of search results pages for certain queries. These summaries synthesise information from multiple sources and present it in a conversational format, complete with citations and related follow-up questions. The feature rolled out globally throughout 2024 after extensive testing under the name Search Generative Experience, fundamentally altering how millions of people interact with search results every day.

The technology behind AI Overviews relies on Google’s large language models, specifically variations of their Gemini architecture. When you submit a query, Google’s systems determine whether your question would benefit from an AI-generated summary rather than traditional search results. If the algorithm decides an overview is appropriate, it generates a response by analysing relevant web pages, synthesising the information and presenting it in paragraph form with clickable source citations. The feature doesn’t replace traditional search results but sits above them, creating a new prime position that websites can appear in through citation rather than direct ranking.

The business implications are staggering. Google processes roughly 8.5 billion searches per day, and AI Overviews now appear for a substantial portion of these queries. Publishers who previously relied on organic search traffic face a uncomfortable reality where users can get their answers without clicking through to the source website. The New York Times, HuffPost and other major publishers have reported measurable traffic declines that correlate with AI Overview deployment. Meanwhile, smaller sites and independent creators find themselves competing not just with other websites but with an AI that can read their content and present the key points directly in the search interface.

Google’s Commercial Incentive to Compete With Their Own Customers

The strategic calculus behind AI Overviews reveals Google’s response to an existential threat. ChatGPT demonstrated that users would happily abandon search engines for conversational AI interfaces that provide direct answers. Within two months of ChatGPT’s launch in November 2022, the tool reached 100 million users. Internal Google memos reportedly described the situation as a “code red” moment. The company faced a genuine possibility that users might start asking questions to ChatGPT instead of Google, particularly for complex queries that traditionally required multiple searches and clicks.

Google’s advertising business depends entirely on users clicking through to websites where ads can be displayed. AI Overviews create an obvious tension where providing better user experience means fewer clicks and potentially less advertising revenue. The company’s solution involves a calculated gamble that keeping users on Google Search, even if they click less, beats losing them entirely to ChatGPT or Perplexity. Early data suggests this bet might pay off. Google has stated that queries with AI Overviews show higher click-through rates to websites than those without, though independent verification of these claims remains limited and several publishers dispute these findings.

The feature also serves as a defensive moat against Microsoft’s Bing, which integrated OpenAI’s technology months before Google launched AI Overviews. Microsoft’s partnership with OpenAI gave them first-mover advantage in AI-powered search, prompting Google to accelerate their own timeline. The competitive pressure was intense enough that Google shipped AI Overviews despite internal concerns about accuracy and reliability. Several embarrassing incidents in the early rollout, including an AI Overview that suggested adding glue to pizza, highlighted the risks of rushing generative AI features to market. Google subsequently refined the systems and reduced the frequency with which overviews appear, particularly for queries involving medical advice, news and other sensitive topics.

The Data Processing Stages for AI Response Generation

Understanding the selection mechanism matters enormously for anyone who publishes content online. Google’s algorithms evaluate whether a webpage can contribute to an AI Overview based on several factors including topical authority and alignment with search intent. The system doesn’t simply scrape the top-ranking pages. Instead, it analyses content from across the web, including pages that might rank on the second or third page of traditional results. This means websites can appear in AI Overviews even if they don’t crack the top ten organic positions, creating new opportunities for visibility.

The citation format in AI Overviews functions differently from traditional search rankings. When Google’s AI references your content, it includes a small clickable citation that links to your page. These citations appear in line with the generated text, similar to how academic papers reference sources. Multiple websites can be cited within a single overview, and the same site might receive several citations if it covers different aspects of the query. Early analysis from SEO researchers suggests that being cited in an AI Overview generates fewer clicks than ranking in position one or two of traditional results, but more clicks than appearing further down the page.

Content structure significantly influences whether Google’s AI will reference your material. Pages with clear headings, well-organised information, authoritative links and correct tone appear more frequently in AI Overviews than those with thin content or excessive advertising. The system seems to prefer pages that directly answer questions without requiring users to scroll past ads or irrelevant information. Google has stated that E-E-A-T principles (Experience, Expertise, Authoritativeness and Trustworthiness) apply to AI Overview selection, meaning the same quality signals that influence traditional rankings also affect whether your content gets cited.

How Site Visitor Retention and Engagement is Changed

Publishers face a mathematics problem with no clear solution. If AI Overviews reduce clicks by even 10-20%, the financial impact becomes severe for sites operating on tight margins. Advertising revenue, affiliate income, donations and subscription conversions all depend on users visiting websites rather than reading summaries on Google. Some publishers report traffic declines exceeding 20% for certain categories of content, particularly informational queries where users seek quick answers rather than in-depth analysis.

Recipe sites provide a telling case study. These sites traditionally ranked well for cooking-related searches, generating revenue from ads displayed alongside recipe instructions. AI Overviews can now present the recipe directly in search results, complete with ingredients and steps, reducing the incentive to click through. The same pattern plays out for how-to guides and factual queries where the answer can be summarised in a few paragraphs. Content that provides unique perspectives, original reporting or detailed analysis appears somewhat more resistant to this effect, but no category of publisher remains entirely immune.

The situation creates perverse incentives that could degrade web content quality over time. If publishers know Google will present their information without clicks, they have less reason to invest in comprehensive, well-researched content. Why spend resources on detailed articles if an AI will extract and present the key points? Some publishers have begun blocking Google’s AI crawlers or hiding content behind paywalls, but these strategies carry their own risks. Blocking Google means losing visibility entirely, while paywalls eliminate the casual traffic that often converts to subscribers. The entire system depends on publishers continuing to create content that Google’s AI can then summarise, creating a tragedy of the commons scenario where individual rational behaviour leads to collective dysfunction.

Technical Methods for Maintaining User Experience Standards

The emerging field of AI Overview optimisation requires rethinking content strategy without abandoning what makes content valuable. Structured information architecture helps both human readers and AI systems parse your content efficiently. Start with clear, direct answers to common questions within the first few paragraphs. Google’s AI seems to favour content that doesn’t require extensive interpretation or inference. If someone searches for “how long does it take to bake salmon,” a page that states “Salmon typically bakes for 12-15 minutes at 200°C” within the opening section stands a better chance of citation than one that buries this information after four paragraphs of preamble.

Depth and specificity matter more than ever. Thin content that merely rehashes information available elsewhere offers little value to Google’s AI or to readers. Original research, unique data, expert interviews and detailed explanations provide citation-worthy material that stands out. Consider creating content that goes beyond surface-level answers to address follow-up questions users might have. If your page about baking salmon also covers how to tell when it’s done, what temperature to check for, how different thicknesses affect cooking time and guides for seasoning, you’ve created multiple citation opportunities within a single piece.

Technical implementation supports content quality but can’t replace it. Use semantic HTML with proper heading hierarchy. Ensure your schema markup accurately describes your content type, whether that’s a recipe or article. Google’s AI systems can interpret structured data and may use it when generating overviews. However, resist the temptation to stuff keywords or manipulate content solely for AI citation. Pages that prioritise algorithmic optimisation over human readability tend to perform poorly in the long term as Google refines its systems to identify and demote content designed primarily for machines.

The Accuracy Problem That Google Can’t Seem to Solve

AI Overviews inherit all the reliability issues of large language models plus additional complications from interpreting web content. The system occasionally misattributes information or combines details from multiple sources in ways that create new falsehoods. During the initial rollout, numerous examples circulated on social media showing AI Overviews providing dangerous or absurd advice. Sometimes the AI cannot distinguish between what is real guidance and what comes from satirical comments from users on sites such as Reddit.

Medical and health queries present particular challenges. Google explicitly reduced AI Overview frequency for these searches after researchers identified multiple instances of incorrect medical information. The problem stems from the fundamental nature of language models, which predict probable text sequences rather than reasoning about truth. When web content contains conflicting information about health topics, the AI might generate a summary that sounds authoritative but contradicts medical consensus. The stakes become considerably higher when incorrect information could influence medical decisions.

Google has implemented various safeguards including reduced coverage for sensitive topics and user feedback mechanisms. The company now generates AI Overviews for fewer queries than during the initial beta period, focusing on cases where the technology performs reliably. Users can report inaccurate overviews through a feedback button, though the effectiveness of this crowdsourced quality control remains unclear. The accuracy challenge reveals a broader tension in generative AI deployment where the technology works impressively most of the time but fails in unpredictable ways that undermine user trust. For Google, whose brand identity centres on providing reliable information, these failures carry reputational risks that extend beyond any individual search result.

What AI Overviews Mean for the Future of Search Engine Optimisation

SEO professionals find themselves recalibrating strategies that evolved over two decades of relatively stable search interfaces. Traditional tactics like keyword optimisation and backlink building remain important but no longer tell the complete story. The new reality involves optimising for two distinct outcomes, appearing in AI Overview citations and ranking well in traditional results. These objectives overlap considerably but not entirely, creating complexity for resource allocation and strategy.

Citation frequency appears to correlate with traditional ranking signals but the relationship isn’t linear. A page ranking in position seven for a query might get cited in the AI Overview while the position three result doesn’t, depending on how well each page answers specific aspects of the query. This means focusing exclusively on ranking position becomes less useful as a metric. Instead, SEO now requires thinking about topical authority across multiple related queries. Building comprehensive content clusters that thoroughly cover a subject area increases the likelihood of citations across numerous AI Overviews, even if no single page dominates traditional rankings.

The shift also affects link building strategy. Links remain a core ranking signal, but their value increasingly comes from topical relevance and authority transfer rather than raw PageRank. Google’s AI systems evaluate content quality partly through the same signals that influence traditional rankings, meaning authoritative backlinks still matter. What’s changing is the endpoint. Instead of trying to rank a single page for a single keyword, modern SEO requires building topical authority that makes your content citeable across many related queries. This demands more sophisticated content planning and a willingness to create material that may never rank traditionally but supports overall domain authority.

How Businesses Should Adapt Content Strategy for AI-Powered Search

Companies that depend on organic search traffic need strategies that account for reduced click-through rates while maintaining content quality. The answer isn’t to stop creating content, but to diversify how that content generates value. Consider content as a multi-purpose asset that drives search visibility by establishing expertise and supporting direct relationships with audiences. A detailed guide might generate fewer clicks from search if Google presents the key points in an AI Overview, but it still demonstrates your expertise to the people who do visit, potentially converting them to email subscribers or customers.

Building owned audiences becomes more important when rented platforms like search become less reliable. Email list subscribers social media followers provide direct access to audiences without algorithmic intermediaries. Content that ranks well in search can serve as a top-of-funnel acquisition channel that feeds these owned platforms. The key is ensuring your content offers enough value that people who find it through AI Overviews still have reason to click through and potentially join your audience directly.

Product and service-focused businesses need different tactics than publishers. If you sell something, create content that addresses user questions but naturally connects to what you offer. A plumbing company writing about common pipe problems wants to be cited in AI Overviews but also wants readers to recognise they need professional help. The solution involves creating genuinely helpful content that establishes credibility while making the case for professional services obvious. AI Overviews might present your content, but they can’t fix a burst pipe, giving users clear reason to click through and contact you.

With nearly 20 years of experience in professional SEO and digital marketing, we understand what it takes to succeed in an AI-powered search environment. Our main hub is based in Horley, Surrey, with additional locations in Peckham and Hampstead in London. Whether you need a content strategy that performs in AI Overviews or comprehensive SEO services that drive results, we can help you adapt with confidence. Get in touch to see how we can bring your digital presence to life with professional expertise.