Google AI Overviews Is Rewriting the SEO Rulebook

Google AI Overviews is changing what it means to rank. Instead of sending users to your site, Google now synthesises answers directly in the search results, pulling from multiple sources and presenting a summary before anyone clicks anything. For marketers who have spent years optimising for position one, this is a structural shift in how organic traffic gets distributed, and it requires a clear-eyed response rather than panic.

The short version: AI Overviews will reduce click-through rates on informational queries, increase the importance of being cited as a source rather than simply ranking, and push low-quality content further into irrelevance. The longer version is more nuanced, and that nuance is where the real strategic opportunity sits.

I spent the early part of my career obsessed with performance marketing metrics. Click-through rates, cost-per-acquisition, position tracking. It felt like the most honest part of the job because the numbers were right there. What I understood later, after running agencies and managing hundreds of millions in ad spend across thirty industries, is that a lot of what performance marketing claims credit for was going to happen anyway. Someone already searching for your brand was already close to buying. Capturing that intent is not the same as creating demand. AI Overviews is exposing that distinction in a way that should make every SEO-heavy marketing team reassess what they are actually building.

If you want the broader context for how AI is reshaping the discipline, the AI Marketing hub covers the full picture, from content creation to search strategy to measurement.

What Are AI Overviews and How Do They Work?

AI Overviews (formerly Search Generative Experience, or SGE) appear at the top of Google search results for a growing range of queries. Google’s large language model synthesises information from multiple web pages and presents a coherent answer, often with expandable citations. The user gets what looks like a direct answer without needing to click through to any source.

The mechanism matters here. Google is not just pulling a featured snippet from one page. It is generating a response, drawing on content from multiple sources, and attributing those sources in a carousel beneath the summary. This means the relationship between ranking and visibility has changed. You can be cited in an AI Overview without being in position one. You can be in position one and not be cited at all.

Google has been careful about where AI Overviews appear. They are more common on informational and how-to queries, less common on navigational searches (where someone is looking for a specific site), and generally absent from transactional queries where commercial intent is clear. That pattern is not accidental. Google needs to protect its advertising revenue, and showing a generative answer above paid search results for high-intent commercial queries would undermine that model.

Understanding the foundational elements of SEO in an AI-driven search environment helps clarify which signals Google is using to decide what gets cited and what gets ignored.

Which Types of Content Are Most Affected?

Not all SEO traffic is equally at risk. The impact depends almost entirely on query intent, and this is where a lot of the breathless commentary about AI Overviews misses the point.

Informational content, the kind that answers “what is”, “how does”, “why does” questions, is most exposed. If someone searches “how does compound interest work”, Google can now answer that directly. The click to a personal finance blog that used to capture that query may not happen. For publishers whose business model depends on high-volume informational traffic monetised through display advertising, this is a serious structural problem.

Commercial and transactional content is less affected, at least for now. Searches like “best project management software for agencies” or “buy running shoes under 100 dollars” still tend to surface traditional results, product listings, and paid ads. Google’s incentive to keep these results clean is strong.

Review and comparison content sits somewhere in between. Google may summarise a comparison in an AI Overview, but users making purchasing decisions often want to read the detail themselves. The click-through rate on these queries may drop, but it is unlikely to collapse entirely.

Local queries are also relatively protected. “Plumber near me” or “best Italian restaurant in Manchester” still need local results, maps, reviews, and real-time data that generative summaries cannot fully replace. If you are doing local SEO, the calculus is different.

The Moz research on AI content offers useful framing on how content quality is being evaluated in this new environment, and it is worth reading if you are trying to understand what “good content” means to Google’s generative layer specifically.

Does Being Cited in an AI Overview Actually Matter?

This is a question I have been thinking about carefully, because the instinct is to say “yes, obviously” and move on. But the honest answer is more complicated.

Being cited in an AI Overview does not reliably drive significant traffic. The citation appears as a small link beneath the generated summary. Many users read the answer and leave. But there is a secondary effect worth considering: brand presence. If your brand name or domain appears consistently in AI-generated answers on topics relevant to your category, that visibility has compounding value even without a click. It signals authority to users who see it repeatedly. It may influence brand recall at the moment of a later commercial search.

I think about this the way I think about reach versus conversion. Early in my career I undervalued the top of the funnel because I could not measure it as cleanly. The performance numbers were right there, and they felt like truth. But the brands that grew consistently were the ones investing in being known, not just in capturing people who were already looking. AI Overview citations are a form of presence, not conversion. Treating them purely as a traffic source misses the point.

The practical implication is that you should be tracking citation presence alongside traditional rank tracking. These are different signals. A page that ranks third but gets cited in every relevant AI Overview may be delivering more brand value than a page sitting in position one for a query that now gets suppressed by a generative answer.

Knowing how an AI search monitoring platform can improve your SEO strategy becomes directly relevant here. Traditional rank trackers were not built to capture this kind of visibility data.

How Do You Optimise Content to Appear in AI Overviews?

There is no confirmed checklist from Google. What we can observe from patterns in which content gets cited is that structure, clarity, and genuine authority matter more than keyword density.

Content that answers a specific question in the first paragraph, uses clear headings, and provides accurate factual information tends to perform better in generative results. This is not a new insight, but AI Overviews have sharpened the penalty for content that buries its answer in filler or wraps a simple point in unnecessary padding.

The Ahrefs AI SEO webinar covers the technical side of this in detail, including how Google’s systems evaluate content structure and what signals appear to correlate with citation frequency.

A few principles that hold up under scrutiny:

• Answer the question directly and early. Google’s generative layer favours content that gets to the point. If your page spends three paragraphs establishing context before answering the question in the title, you are at a disadvantage.
• Use structured formatting. Lists, tables, and clear subheadings help Google’s systems parse your content. This is not about gaming the algorithm; it is about making your content easier to understand, which has always been good practice.
• Build topical depth, not just breadth. A single comprehensive page on a specific topic tends to outperform a cluster of thin pages. Google’s generative layer appears to favour sources that demonstrate genuine expertise on a subject, not pages that mention a keyword enough times.
• Earn citations and links from authoritative sources. The content that gets cited in AI Overviews tends to come from domains with established authority. This is not a shortcut you can manufacture. It requires doing the work of being genuinely useful and credible over time.

Understanding how to create AI-friendly content that earns featured snippets is a practical starting point for teams who want to adapt their content production to this environment.

What Does This Mean for Content Strategy More Broadly?

The brands I have seen struggle most with shifts like this are the ones that built their SEO strategy around volume. Hundreds of informational articles targeting long-tail keywords, optimised to rank and capture passive search traffic. That model worked for a long time. It is now under pressure from multiple directions simultaneously: AI Overviews reducing click-through rates, Google’s helpful content updates penalising low-quality content, and user behaviour shifting toward more conversational search patterns.

The brands in a stronger position are the ones that invested in building genuine authority on specific topics, creating content their audience actually finds useful, and developing a recognisable point of view. These are the brands that get cited, not just ranked.

There is an analogy I come back to from retail. A customer who tries something on in a shop is far more likely to buy than one who just browses. The act of engagement changes the probability. Content that genuinely engages your audience, that makes them think, that gives them something they did not have before, does the same thing. It creates preference, not just awareness. AI Overviews cannot replicate that. They can summarise facts. They cannot build a relationship or establish a distinctive voice.

This is why the brands that will weather this shift most effectively are the ones investing in distinctive, authoritative content rather than optimised commodity content. The efficiency gains from AI-powered content creation are real, but efficiency applied to the wrong strategy just produces more of the wrong thing faster.

For teams thinking about how to operationalise content production in this environment, the SEO AI agent content outline framework is worth reviewing as a structural starting point.

How Should You Measure SEO Performance Now?

This is where most teams are currently underprepared. Traditional SEO measurement, rank position, organic traffic, click-through rate, was built for a world where ranking meant clicking. That relationship is weakening.

The measurement framework needs to expand. Alongside traditional metrics, you should be tracking:

• AI Overview citation frequency: How often does your content appear as a cited source in AI Overviews for your target queries?
• Branded search volume: If AI Overviews are building awareness without clicks, one downstream effect should be an increase in branded searches over time. This is imperfect but worth watching.
• Direct and return traffic: Users who encounter your brand in an AI Overview and later visit directly are not captured in organic click-through data. Tracking these patterns gives a more complete picture.
• Conversion rate on organic traffic: If overall organic volume drops but the traffic that does arrive is more commercially qualified, your conversion rate should improve. A drop in traffic that comes with a rise in conversion rate is not necessarily a bad outcome.

The Semrush overview of AI optimisation tools covers some of the emerging measurement options, though this space is moving quickly and any specific tool recommendation needs regular re-evaluation.

I judged the Effie Awards for several years, and one thing that process reinforced is that the most credible effectiveness cases are built on multiple signals pointing in the same direction, not single metrics. The same discipline applies here. No single number tells you how AI Overviews are affecting your brand. You need a set of signals that, taken together, give you an honest picture.

If you want a broader framework for thinking about AI’s role in marketing measurement and strategy, the AI Marketing Glossary is a useful reference for getting the terminology right before you start evaluating tools or briefing teams.

Is This the End of SEO?

No. But it is the end of a particular version of SEO that was always more fragile than it appeared.

The version of SEO that involved producing large volumes of informational content to capture passive search traffic was always dependent on Google’s goodwill. Every major algorithm update over the past decade has moved in the same direction: reducing the effectiveness of content that exists primarily to rank rather than to genuinely serve the reader. AI Overviews are the most significant expression of that direction yet, but they are not a departure from it.

What survives is what always should have been the priority: content that demonstrates real expertise, earns genuine authority, and serves a specific audience better than the alternatives. The Ahrefs perspective on AI tools in SEO makes a similar point, that the fundamentals of good SEO have not changed, even if the tactics around them need to adapt.

What this shift does is accelerate the divergence between brands that invested in genuine authority and brands that invested in volume. If you are in the first group, AI Overviews are an opportunity to be cited, to be visible, to be the source that Google’s generative layer trusts. If you are in the second group, the traffic erosion that may have been gradual is likely to become faster.

The practical question is not whether to do SEO. It is whether your SEO strategy is oriented toward building something real or toward capturing something that was always going to be temporary.

There is more on the evolving relationship between AI and marketing strategy across the AI Marketing section of The Marketing Juice, including how brands are adapting content, measurement, and channel strategy to this new environment.

About the Author

Keith Lacy is a marketing strategist and former agency CEO with 20+ years of experience across agency leadership, performance marketing, and commercial strategy. He writes The Marketing Juice to cut through the noise and share what works.

Frequently Asked Questions