Reviewed by the SEOPointz team · Last reviewed June 2026. We’ve updated this piece to reflect how the 2025 predictions actually played out, using published CTR and AI Overview studies rather than forecasts. SEOPointz may earn a commission from some links; it never changes what we recommend.
When everyone was writing “SEO trends for 2025” lists at the start of the year, most of them hedged. The honest version, looking back from mid-2026, is simpler: 2025 was the year AI search stopped being a side experiment and became the surface most of your audience actually sees first. If you want to stay ahead of the game now, the question isn’t “which keyword tactic is hot” — it’s “how do I stay visible when Google answers the question before anyone reaches my site.” Here’s what genuinely changed, and what to do about it.
AI Overviews went from novelty to default
This is the trend that reshaped everything else. According to Semrush’s 2025 study, AI Overviews appeared on roughly 13% of U.S. desktop queries by March 2025 — about double their reach two months earlier. The feature expanded fastest on informational and how-to queries, exactly the kind of content most blogs are built around. By 2026, treating AI Overviews as an edge case is no longer defensible; for many topics they are the first thing a searcher reads. The practical takeaway is that ranking #1 in the classic blue links and being cited inside the AI answer are now two separate battles, and the second one increasingly decides whether you get the click.
The click-through math got harder — but not hopeless
The numbers are sobering. An Ahrefs analysis from December 2025 found that when an AI Overview is present, the top-ranking result loses roughly 58% of its expected clicks — up from a 34.5% drop measured back in April. Seer Interactive reported organic CTR on informational queries falling about 61% since mid-2024. So yes, the same ranking earns fewer visitors than it did two years ago.
The nuance worth keeping: pages cited inside an AI Overview tend to earn more clicks than uncited pages on the same results page. And there are early signs of stabilization — Seer tracked the CTR on AI Overviews themselves recovering from a low of 1.3% in December 2025 to 2.4% by February 2026. The lesson isn’t “traffic is dead.” It’s that being the source an AI quotes is now the high-value position.
AEO and GEO became real disciplines
Two acronyms went from jargon to job descriptions in 2025. Answer Engine Optimization (AEO) focuses on being the clean, direct answer that featured snippets and voice assistants read aloud. Generative Engine Optimization (GEO) focuses on being cited by generative tools like ChatGPT, Perplexity and Google’s AI Mode. They overlap with classic SEO — crawlable pages, good information architecture, internal links — but they reward a slightly different style: a clear thesis up front, specific supporting evidence, and a conclusion an AI can lift with confidence. If your content buries the answer three scrolls down under keyword-stuffed preamble, it’s hard to cite. Structure for “dual readability”: clear for humans, extractable for machines.
E-E-A-T pushed past YMYL into nearly everything
Experience, Expertise, Authoritativeness and Trustworthiness used to be a concern mostly for “your money or your life” topics like health and finance. In 2025 that widened. As AI-generated content flooded the web, Google leaned harder on signals that separate genuine expertise from scaled output: named authors with real credentials, original data, first-hand testing, and transparent sourcing. Content that simply re-summarizes what’s already ranking gets filtered faster and recovers slower after core updates. The practical move is unglamorous but effective — put a real byline on your work, show the testing or research behind a claim, and cite your sources the way this article does.
Google AI Mode opened a second front
Late in the year Google rolled out AI Mode, a more conversational, multi-step search experience that goes further than the inline AI Overview. What makes it strategically interesting is that its citations don’t mirror the old results — one analysis found only about 14% of URLs cited in AI Mode also appeared in AI Overview citations for the same queries. In plain terms, AI Mode is a distinct visibility opportunity, not a copy of page one. Brands that earn citations there are often ones with strong topical depth and clear, well-structured pages rather than just high domain authority.
What to actually do in 2026
Synthesizing all of it: keep doing fundamentals well, but reweight your effort. Build genuinely original content — case studies, first-hand reviews, and your own data — because that’s what AI systems prefer to cite and what core updates reward. Structure pages so the answer is easy to extract. Strengthen author and trust signals across your whole site, not just sensitive topics. And measure success on visibility and citations across Google, ChatGPT and Perplexity, not on raw blue-link rankings alone, since Gartner projected traditional search volume could fall around 25% by 2026. The brands staying ahead aren’t fighting AI search — they’re making themselves the thing it quotes.
| Trend | What changed in 2025 | Your move |
|---|---|---|
| AI Overviews | Reached ~13% of U.S. desktop queries by March; top-result CTR down ~58% when present | Optimize to be cited inside the answer, not just rank below it |
| AEO / GEO | Emerged as distinct disciplines for snippets and AI citations | Lead with a clear answer; structure for dual readability |
| E-E-A-T | Extended well beyond YMYL topics | Real authors, original data, transparent sourcing |
| AI Mode | Launched with citations only ~14% overlapping classic results | Build topical depth; treat it as a separate channel |
Frequently asked questions
Is SEO dead because of AI Overviews?
No, but it has shifted. Clicks per ranking are down — the top result can lose well over half its clicks when an AI Overview appears — yet the pages quoted inside those answers still earn meaningful traffic. The goal moved from “rank first” to “be the cited source.”
Do I need to optimize for ChatGPT and Perplexity separately?
Increasingly, yes. The fundamentals overlap with Google SEO, but generative engines favor content with a clear answer, credible sourcing and structured formatting. With roughly a third of U.S. users expected to use generative AI search in 2026, ignoring those surfaces leaves visibility on the table.
Does AI-written content still rank?
It can, but thin AI content that only re-summarizes existing pages tends to get filtered out faster after core updates. What wins is original analysis, first-hand experience and real data — whether a human or an AI assist helped produce the draft.
Want to go deeper? Pair this with our look at SEO best practices to stay ahead of the curve, and for the bigger picture see the evolution of SEO from past to present.
