Google AI Overviews Are Changing SEO. Here Is What Shifted
Google AI Overviews have changed the search results page in ways that matter commercially, not just cosmetically. When a summarised answer appears above organic results, the click that previously went to the ranking page often disappears entirely. The question worth asking is not whether AI Overviews are significant, but which types of content are losing traffic, which are holding, and what that means for how you allocate SEO resource going forward.
The short answer: informational queries are taking the biggest hit, transactional and navigational intent is largely intact, and the brands best positioned are those that built authority around depth rather than volume. Everything else is noise.
Key Takeaways
- AI Overviews appear most frequently on informational queries, which means content built purely to capture top-of-funnel traffic is the most exposed to click loss.
- Transactional and navigational queries are largely unaffected, which reinforces the case for SEO programmes built around commercial intent rather than raw traffic volume.
- Being cited inside an AI Overview is a real visibility signal, and it correlates with the same factors that drive traditional rankings: depth, authority, and clear sourcing.
- Measuring SEO performance by clicks alone has always been a flawed proxy. AI Overviews have made that flaw impossible to ignore.
- The brands that will fare best are those that treat SEO as an authority-building programme, not a content production treadmill.
In This Article
What Are Google AI Overviews and Why Do They Change the Calculus?
Google AI Overviews, rolled out broadly in the US in May 2024 and expanding internationally since, are AI-generated summaries that appear at the top of search results for a wide range of queries. They pull from multiple sources, synthesise an answer, and present it before the user ever scrolls to an organic listing. For some queries, the answer is complete enough that clicking through to a source becomes optional.
This is not a small tweak to the results page. It is a structural change to how Google monetises attention and how organic content competes for clicks. The SEO industry has spent two decades optimising for position one. Position one still matters, but what sits above it now is a summarised answer that may have absorbed the click entirely.
I have spent enough time managing large-scale paid and organic search programmes to know that the instinct in moments like this is to panic and over-rotate. I saw it when featured snippets arrived, when People Also Ask boxes proliferated, and when Google Shopping reshaped retail search. Each time, the businesses that responded with clear thinking rather than reactive output came out ahead. The same principle applies here.
If you want to understand how AI Overviews fit into a broader view of search strategy, the Complete SEO Strategy hub covers the full picture, from technical foundations to content and authority building.
Which Queries Are Most Affected?
The pattern that has emerged through 2024 is reasonably consistent. AI Overviews appear most often on informational queries: definitions, how-to questions, explanatory searches, and general knowledge lookups. These are precisely the query types that content teams have been targeting for years under the logic that top-of-funnel traffic builds brand awareness and feeds the conversion pipeline downstream.
That logic was always shakier than most SEO teams admitted. When I was running agency programmes across thirty-plus industries, we would regularly audit organic traffic and find that enormous volumes of informational traffic were contributing almost nothing measurable to revenue. The attribution models looked impressive. The commercial reality was often more modest. AI Overviews have not created that problem. They have exposed it.
Transactional queries, those with clear purchase or comparison intent, are largely unaffected. Google has a financial incentive to keep those results clickable, because they feed the paid auction. Navigational queries, where someone is looking for a specific brand or site, are also stable. The exposure is concentrated in the informational middle ground, and that is where many content-heavy SEO programmes have been over-invested for years.
Semrush’s analysis of Google AI Mode’s SEO impact offers a useful perspective on how these shifts are playing out across different query categories. The picture it paints is consistent with what I have observed in practice: the disruption is real but uneven, and intent-based segmentation matters more than ever.
Does Being Cited in an AI Overview Matter?
Yes, and more than most teams currently treat it. When Google’s AI Overview cites a source, it typically displays a link. Those citations are not generating the same click volumes as a traditional ranking, but they are a visibility signal and a brand credibility signal that should not be dismissed.
The question is what drives citation. The answer, based on what has been observed across the industry through 2024, is that it correlates strongly with the same factors that drive traditional rankings: topical authority, content depth, clear sourcing, and structured information. Content that is thin, generic, or written primarily to hit keyword density is not being cited. Content that demonstrates genuine expertise and provides clear, verifiable information is.
Moz’s thinking on SEO and content predictions for 2024 flagged this shift early: the move toward demonstrable expertise is not a trend, it is the direction of travel for the entire discipline. That is consistent with what Google has been signalling through E-E-A-T guidance for several years. AI Overviews have accelerated the practical consequences of ignoring that signal.
There is also a longer-term brand consideration. If your content is being cited inside AI Overviews regularly, you are building familiarity with users who may never click through on that session but will recognise your brand when they encounter it later. That is not easily measurable, which is precisely why most performance-focused teams will undervalue it. I have judged Effie Award entries where the winning work was built on exactly that kind of hard-to-attribute brand presence. The measurement challenge does not make the effect less real.
What Has Actually Happened to Organic Click-Through Rates?
The honest answer is that the data is messy and context-dependent. Click-through rates for queries where AI Overviews appear have declined in aggregate, but the degree varies significantly by query type, industry, and whether the AI Overview answer is complete or partial.
For queries where the AI Overview provides a full answer, clicks to organic results drop materially. For queries where the overview is partial or prompts further investigation, the impact is smaller. The practical implication is that content designed to answer a question completely in a single paragraph is more exposed than content that covers a topic with enough depth that the AI Overview becomes an entry point rather than a destination.
This is where I would push back on the instinct to simply write longer content. Length alone does not protect you. I have reviewed content programmes that produced three-thousand-word articles that said nothing a five-hundred-word piece could not have covered. The issue is depth and differentiation, not word count. Proprietary data, original analysis, practitioner perspective, and genuinely useful frameworks are what make content worth citing and worth clicking through to read in full.
Moz’s work on AI for SEO and content marketing makes a related point: AI-assisted content production has lowered the cost of creating average content dramatically, which means average content is now worth less than it has ever been. The floor has dropped. The ceiling, for genuinely differentiated content, has not.
How Should SEO Strategy Adapt?
There are four practical adjustments worth making, none of which require abandoning what works.
First, audit your informational content by commercial proximity. Not all informational traffic is equal. Content that sits close to a purchase decision, comparison guides, category explainers, use case breakdowns, has more commercial value than content that answers general questions with no clear path to conversion. Prioritise the former and be honest about the latter.
Second, invest in content that cannot be easily summarised. Original research, proprietary data, case studies, and practitioner perspectives are harder for AI Overviews to synthesise into a complete answer because they contain information that does not exist elsewhere. This is not a new principle. It is a principle that now has sharper commercial consequences if ignored.
Third, structure content for citation. Clear headings, concise definitions, well-organised supporting evidence, and explicit sourcing all increase the likelihood that Google’s systems will identify your content as a credible reference. This is not gaming the algorithm. It is making your expertise legible to both humans and machines.
Fourth, fix your measurement before drawing conclusions. This is the one I feel most strongly about. I have spent years watching organisations make strategic pivots based on flawed measurement, and the arrival of AI Overviews has created a new wave of that problem. Teams are seeing traffic decline and attributing it entirely to AI Overviews when the reality is more complicated. Seasonality, algorithm updates, technical issues, and content decay are all contributing factors. If you cannot isolate the AI Overview effect with reasonable confidence, you should not be making large strategic bets based on it.
There is a broader point here about how the SEO industry measures itself. Organic clicks have always been a proxy for value, not value itself. The businesses I have seen manage SEO most effectively are those that connect organic performance to revenue outcomes, not those that optimise for traffic in isolation. AI Overviews have made the gap between traffic and value more visible. That is uncomfortable, but it is useful.
The Measurement Problem Is the Real Story
When I was at lastminute.com, we ran a paid search campaign for a music festival that generated six figures of revenue within roughly a day. Simple campaign, clear intent, measurable outcome. The feedback loop was immediate and unambiguous. SEO rarely works like that, and AI Overviews have made the feedback loop even noisier.
The temptation is to respond to that noise by measuring more: more metrics, more dashboards, more attribution models. In my experience, that usually makes things worse. More data without better interpretation is just more noise. What the AI Overviews moment actually requires is cleaner thinking about what you are trying to achieve and whether organic search is the right channel to achieve it.
If your SEO programme exists to drive brand visibility among audiences who are not yet ready to buy, AI Overviews may be reducing clicks but not necessarily reducing value, because visibility inside an overview still carries brand exposure. If your programme exists to drive qualified traffic that converts, the question is whether the queries you are targeting still deliver that, and if not, where you should redirect resource.
The foundational principles of SEO that Search Engine Land outlined years ago remain relevant here: the mistakes that damage organic performance are mostly about misaligned priorities and poor fundamentals, not about failing to chase the latest feature. AI Overviews do not change that. They amplify it.
For a more complete view of how these changes fit into a broader SEO programme, including how to balance technical health, content strategy, and authority building, the Complete SEO Strategy hub is worth working through in full.
What to Watch in 2025
AI Overviews are not a finished product. Google has already adjusted how and when they appear following the early rollout issues in 2024, when the feature surfaced some memorably poor answers and generated significant press coverage. The trajectory is toward more AI-generated content in search results, not less, but the specific implementation will continue to evolve.
The queries where AI Overviews appear are likely to expand. Google has a strong incentive to keep users inside its ecosystem, and AI-generated summaries serve that goal. The categories that are currently protected, transactional and navigational queries, may not remain protected indefinitely, particularly as Google tests AI-assisted shopping and comparison features.
Brand authority will matter more, not less. When users see your brand cited repeatedly inside AI Overviews, that creates a familiarity effect that compounds over time. It is not easily measured in a monthly SEO report, but it is real. The organisations that will find this hardest to accept are those with short reporting cycles and narrow definitions of what counts as a return from SEO investment.
The broader strategic question is whether search, as a channel, continues to deliver the same commercial returns as it has over the past decade. For some categories and query types, the answer is clearly yes. For others, particularly those where informational content has been the primary vehicle, the honest answer is that the channel economics are changing and the resource allocation should reflect that.
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.
