Key Takeaways
- The how will ai affect seo isn’t one change it’s several happening together: AI Overviews reducing clicks, ranking algorithms judging content differently, and citation replacing position as the real target.
- AI Overviews now appear on roughly 47–64% of search queries, up from 25–30% in 2024, and can cut click-through rates on top organic positions by 30–50%.
- Getting cited inside an AI Overview can earn more clicks than holding a traditional top ranking alone visibility and ranking position are no longer the same thing.
- Google’s AI systems RankBrain, BERT, MUM, and their successors increasingly judge content on depth, originality, and trustworthiness rather than keyword matching.
- This shift doesn’t replace SEO. It adds a related discipline, often called AI SEO or generative engine optimization (GEO), that most businesses haven’t started addressing yet.
AI Overviews Are Already Cutting Into Your Click-Through Rate
AI Overviews the AI-generated summaries that now appear above traditional search results aren’t a minor test feature anymore. They show up on roughly 47–64% of search queries today, up from 25–30% when the feature first rolled out widely in 2024.
This continues a pattern search has been building toward for years Google has been showing more answers directly on the results page, without a click, since long before AI Overviews existed: featured snippets, knowledge panels, direct unit conversions. AI Overviews are the same “zero-click” pattern applied at a much larger scale, across far more query types than a snippet ever could cover.
That expansion has a direct, measurable effect: queries with an AI Overview present see an average click-through rate drop of 30–50% on the pages that would have otherwise ranked at the top. Ranking #1 no longer guarantees the traffic it used to.
Not every query is affected equally. Broad informational questions the kind with a short, factual answer are far more likely to trigger an AI Overview than specific commercial or transactional searches, where Google still leans on traditional results. If your traffic comes mostly from “how to” and “what is” style queries, you’re more exposed to this shift than a business whose traffic comes from branded or highly specific searches.
There’s a flip side worth knowing before this feels purely negative. Getting cited inside an AI Overview can earn more clicks than holding a traditional top ranking alone the visibility just comes from a different place on the page than it used to. Google surfaces the AI Overview above the organic results, but it also shows which sources it pulled from, and a citation there gets seen by nearly everyone who looks at the page at all.
For your business, this means checking Google Search Console for queries where impressions have stayed flat but clicks have dropped is now worth doing specifically that pattern often points to AI Overviews appearing on queries you already rank well for.
The Ranking Algorithm Itself Now Thinks More Like AI
This is the part of how AI is changing SEO that’s easiest to miss, because it happened gradually across several updates rather than in one visible event. Google’s ranking systems Rank Brain, BERT, MUM, and the systems that followed them use AI to understand what a searcher actually means, not just which words they typed.
Each one solved a different piece of the same problem:
- RankBrain (2015) helped Google interpret ambiguous or never-seen-before queries by relating them to concepts it already understood, rather than requiring an exact keyword match.
- BERT (2019) improved how the algorithm reads context the difference between “bank” as a financial institution and “bank” as a riverbank, or how a small word like “to” can flip a query’s meaning entirely.
- MUM went further still, connecting information across languages and content formats to answer more complex, multi-part questions in a single pass instead of requiring several separate searches.
In practice, that shifted what “good content” looks like to the algorithm. Depth of coverage, originality, and readability now carry more weight than keyword density ever did. Content that reads as trustworthy and current tends to hold its position; content that’s thin, outdated, or inconsistent tends to lose it, even when the target keyword is technically present.
This is the same underlying shift that produced E-E-A-T as a quality framework Google’s systems got better at telling the difference between content written to rank and content written to actually help someone, and started rewarding the second kind more consistently. A page stuffed with a keyword ten times but saying little of substance now competes poorly against a shorter page that actually answers the question well.
None of this happened overnight, which is part of why it’s easy to miss as a single “impact.” A ranking that slips 5% after one update and another 5% after the next doesn’t look like a crisis in the moment. Looked at over a year or two, though, it adds up to a real shift in which pages hold their position and which ones quietly slide and it’s rarely obvious which category a given page falls into without checking.
For your business, this means how AI changes search rankings isn’t really about a new checklist of tactics. It’s about whether your content would hold up if a knowledgeable person actually read it, not just an algorithm scanning for a phrase.
Getting Cited Matters More Than Getting Ranked
Traditional SEO measured success by position did your page land in the top three, top ten, page one. That measurement still matters, but it’s no longer the only one that determines whether someone finds your business, and it’s worth being direct about why: a ranking report can look identical month over month while the actual number of people reaching your site from that ranking quietly declines, simply because fewer of them are clicking through past an AI-generated answer.
AI systems like ChatGPT, Perplexity, and Google’s own AI Overviews build answers by pulling from and citing specific sources. Rather than reading an entire page, they typically retrieve a specific chunk of text that directly answers the question at hand, then cite where it came from. A page can be well-optimized for traditional ranking and still never get pulled into one of these answers, because ranking and citation are judged by overlapping but different criteria extractability and clarity matter more to a citation than they do to a ranking position.
This is also where search is expanding beyond text. Images, video, and audio content are increasingly indexable and citable too, which means SEO in the AI era covers more formats than the page-and-keyword model most businesses are still working from. A product demo video or a well-labeled infographic can now earn visibility that a text-only page competing on the same keyword might not.
The terminology is still settling some people call this AI SEO, others call it AEO (answer engine optimization) or GEO (generative engine optimization), and the exact labels vary by who’s writing about it. What matters more than the name is the underlying goal all of them describe: being the source an AI system reaches for when it builds an answer, not just the page a person happens to click.
For your business, this means the practical goal is shifting from “rank higher” to “be the source that gets cited” a related goal, but not an identical one, and one that requires its own approach. That approach has a name: generative engine optimization, or GEO, and it’s fast becoming as central to a search strategy as SEO has always been.
What This Actually Means for Your Website
None of this makes your existing SEO work worthless. The technical fundamentals crawlability, page speed, clean structure still matter to both traditional rankings and AI citation, and a site that’s technically broken loses on both fronts at once.
What’s changed is that SEO and AI and search engine optimization are no longer quite the same discipline, even though they overlap heavily. AI SEO (or GEO) adds specific, additional work on top of classic SEO content structured for extraction, authority signals AI systems weigh differently than backlinks alone, and a recency standard that’s stricter than most content calendars currently plan for.
This also changes what “finished” looks like for a piece of content. A page optimized purely for traditional SEO could sit unchanged for a year with little consequence, as long as it kept ranking. A page competing for AI citation needs a more active maintenance rhythm, since AI systems weigh how current a source appears to be more heavily than search rankings ever did.
It also changes where effort should go first. Businesses that treat this as an entirely new project a separate team, a separate budget, a from-scratch content library often move slower than businesses that treat it as an extension of what they’re already doing. The content, the technical foundation, and the subject-matter expertise mostly already exist; what’s missing is usually the structural and citation-focused layer on top of it, not a rebuild from zero.
Whoever currently owns SEO in your business is usually the right person to own this too, at least at first. The skill set overlaps heavily understanding search intent, structuring content clearly, building genuine authority even though a few of the specific tactics are new. Treating it as a completely separate specialty before you’ve even audited your current gap tends to slow teams down rather than speed them up.
For your business, this means the question isn’t whether to abandon SEO for something new. It’s whether your current SEO work is missing the additional layer that determines whether AI systems find, trust, and cite what you’ve already built.
What to Check First
Before assuming AI is or isn’t affecting your traffic, three checks will tell you directly:
- Check Search Console for queries where impressions have held steady but clicks have declined. That pattern is the AI Overview CTR effect showing up in your own data, not a ranking problem.
- Search a handful of your core topics in ChatGPT or Perplexity and see whether your brand shows up anywhere in the answer. If it doesn’t, and a competitor does, that’s a citation gap worth investigating.
- Check whether your highest-traffic pages would still read as authoritative and current to a knowledgeable person — not just to a keyword scanner running through an old checklist.
Once you know how this is actually showing up for your business, the next question is what to do about it. Generative Engine Optimization (GEO) Strategies for Brands walks through the specific strategies for closing the gap this post just described.