Every week someone sends me a new headline: “SEO is dead,” “AI agents will replace websites,” “Google traffic is gone.” Business owners ask me which of these they should actually worry about. It’s a fair question, because the noise is loud.
Ahrefs’ Louise Linehan put it well in her July 2026 roundup of AI search trends:
“AI search is the most consequential thing to happen to our industry in a decade… and the most over-narrated.”
— Ahrefs, 5 AI Search Trends I’m Seeing in 2026
So in this article, I want to separate the trends that are backed by data from the ones that are mostly talk. I focus only on AI search: assistant usage, clicks, ads, agents, citation sources, brand mentions and measurement. For the broader picture, including core updates and classic ranking factors, see my guide to SEO trends. At the end, I give you a short list of what I would prioritize next quarter.
Assistants are becoming a default starting point for research
The scale is no longer small. TechCrunch reported on February 27, 2026 that ChatGPT had reached 900 million weekly active users, up from 800 million in October 2025, with 50 million paying subscribers.
But here is a detail many people miss. A big share of AI answers don’t touch the web at all. In Ahrefs’ study of 66.6 million free, logged-out ChatGPT responses from August and September 2026, only 14.4% triggered a web search. The rest came from what the model already knew.
What this means for you, in plain words:
- Your brand needs to be known to the model, not only findable in search. That comes from being widely mentioned across the web over time.
- When the assistant does search, it searches for specific things: current prices, comparisons, reviews and “official” facts. I explain this in detail in my post on how ChatGPT turns one prompt into many web searches.
- Definition-style content is losing value. Ahrefs found definition-style fan-out queries fell by about a fifth in one month. The model doesn’t need to look those up.
Fewer clicks, more influence
This is the trend with the strongest evidence. Ahrefs updated its click study in February 2026, comparing December 2023 with December 2025 across 300,000 keywords. The result was stark:
“For every 100 clicks you could historically earn for a top-ranking page, Google now “keeps” 58.”
— Ahrefs, AI Overviews reduce clicks (2026 update)
In that study, position-one click-through rate on AI Overview keywords dropped from a predicted 3.7% to an actual 1.6%. My post on AI Overviews click loss covers what to do about this keyword by keyword.
Example: what a 58% click loss looks like for one business
Let me put that into numbers with a hypothetical example. Say a home services company ranks first for keywords that bring 10,000 searches a month, and half of those keywords now show an AI Overview.
- Without AI Overviews, at a 3.7% click-through rate, those 10,000 searches would bring about 370 clicks.
- For the 5,000 searches that now show an AI Overview, a 1.6% rate gives about 80 clicks instead of 185.
- So the site gets roughly 265 clicks instead of 370, a drop of about 28% overall, with rankings unchanged.
Nothing is “broken” on that site. The searchers are still reading about the topic; they are just reading the answer on Google. That is why I push clients to measure branded demand and leads, not only clicks.
Ahrefs’ trends piece describes the industry mood shifting from panic to planning. A few developments stand out:
- Google launched a generative AI performance report in Search Console in June 2026. Ahrefs notes its main limitation: it shows impressions in AI Overviews and AI Mode, but not clicks or queries.
- Google added a Search Console control to opt out of AI search features, which Ahrefs links to pressure from the UK’s competition regulator. Critics, as Ahrefs notes, point out that sites get the opt-out but not the data to judge it.
- Legal pressure is growing. Ahrefs highlights a Munich court ruling that treated AI Overviews as Google’s own statements, making Google liable for false answers.
The practical lesson: don’t judge AI search only by clicks. Influence shows up in branded search, direct visits and sales conversations. I explain how to measure that in my guide to proving AI search impact with holdout and pre/post tests.
Ads are arriving in AI answers
The “ad-free answer” era is ending. PPC Land reported on January 16, 2026 that OpenAI confirmed it would test ads in ChatGPT for logged-in adults in the US on the Free and Go tiers. According to that report:
- Ads appear at the bottom of answers and are labeled as sponsored.
- OpenAI says ads do not influence the answers ChatGPT gives.
- No ads appear near sensitive topics like health, mental health or politics, or for users under 18.
- Plus, Pro, Business and Enterprise plans stay ad-free.
Google’s side of this story is different, and I covered both platforms in my guides to ads in ChatGPT and ads in AI Overviews.
What I tell clients: paid placement in AI answers will become another media channel, but organic visibility in the answer itself still matters most. An ad under an answer that recommends your competitor is a hard sell. Our PPC management team treats AI ad inventory as a test budget for now, not a core channel.
Agents and AI browsers are starting to do the clicking
This is the trend with the most hype and the least settled practice. Still, some signals are hard to ignore.
- Bot traffic overtook human traffic. Ahrefs cites Cloudflare CEO Matthew Prince saying in June 2026 that bots, crawlers and agents had passed 50% of internet traffic for the first time.
- Search demand for the topic exploded. Ahrefs reports “agentic seo” searches grew by more than 5,800%, from a very small base.
- Google is building tools for it. Ahrefs notes that Chrome announced an experimental “Agentic Browsing” category at Google I/O, with agent readiness checks in Lighthouse.
- Some popular fixes don’t work yet. Ahrefs’ own study found 97% of llms.txt files are never read.
Google’s own guidance on AI features pushes back against over-engineering:
“There's no requirement to break your content into tiny pieces for AI to better understand it.”
— Google Search Central, Guidance on generative AI features in Search
My advice: make your site easy for agents to use with the basics. That means clean HTML, accessible forms, clear prices and policies in text, and fast pages. Those things help humans too. My post on AI browsers and SEO has a checklist.
Citation sources are shifting fast
For two years, the advice was “be on Reddit and Wikipedia, because AI cites them.” The latest data shows how quickly that can change. Ahrefs’ fan-out study tracked US ChatGPT citations from March to September 2026:
| Source | Change reported by Ahrefs (Mar–Sep 2026) |
|---|---|
| From about 1 in 5 answers to about 1 in 43 (−88%) | |
| Wikipedia | −80% |
| YouTube | −27% |
| PubMed | Rose from 64th to 4th most-cited domain |
| OpenAI publisher partners (share of top 200 domains) | 7.5% in March, 15.8% in July, 12.5% in late September |
Ahrefs also found ChatGPT leaning on official sources: in a test of 20,000 US prompts, 84.4% of fact-based answers cited at least one official site. And 6.8% of all fan-out queries were site: searches, mostly aimed at brand websites.
Two lessons. First, don’t bet everything on one platform. A source that drives citations today can fade in months. Second, your own website is becoming a primary source for facts about your business. Keep prices, specs and policies clear and current. For the full picture by platform, see my guide to the sites ChatGPT cites most.
Brand mentions are outweighing links
Links still matter for Google rankings. But for AI visibility, mentions appear to matter more. Ahrefs’ study of about 75,000 brands, summarized in its guide to measuring AI search visibility, found these correlations with AI mentions:
| Signal | Correlation with AI mentions |
|---|---|
| Branded web mentions (ChatGPT) | 0.656 |
| Branded web mentions (AI Mode) | 0.709 |
| Branded web mentions (AI Overviews) | 0.664 |
| YouTube mentions | About 0.74 |
| Domain Rating, referring domains, backlinks, URL Rating | All below 0.33 |
Correlation is not proof, and I treat this as direction, not hard science. But it fits what I see: brands that get talked about get recommended. That means PR, reviews, podcasts, YouTube and community presence are now part of search work. My guide to AI mentions without links goes deeper.
AI search measurement is finally maturing
A year ago, most teams had no reliable AI data at all. That has changed:
- GA4 has an AI Assistant default channel. Google’s documentation says it covers traffic from AI chatbots such as ChatGPT, Gemini and Copilot, and excludes AI Overviews and AI Mode.
- Search Console reports AI impressions. Google’s AI features documentation says AI Overviews and AI Mode traffic counts in the Performance report under the “Web” search type, and the new generative AI report adds impression detail.
- Bing goes further on queries. Ahrefs notes that Bing Webmaster Tools shows the queries and citation share behind its AI answers.
- Demand for tracking tools is rising. Ahrefs reports searches for “ai search tracking” grew 184% and “ai rank tracking” 175% year over year.
The catch is that AI answers vary a lot from run to run. A SparkToro and Gumshoe study reported by Search Engine Land found the odds of seeing the same brand list twice were under 1 in 100. So the maturing skill isn’t just collecting data; it is deciding what to do with it. That is the focus of my framework for acting on AI visibility data.
A quick summary of what is solid and what is still hype:
| Trend | Evidence strength | My take |
|---|---|---|
| Fewer clicks from AI answers | Strong (large studies) | Plan for it now |
| Assistants as research starting point | Strong (usage data) | Build brand awareness, not just pages |
| Brand mentions over links for AI | Moderate (correlation) | Invest in PR and reviews |
| Citation source shifts | Strong but fast-moving | Diversify; strengthen your own site |
| Ads in AI answers | Early | Test with small budgets |
| Agents doing the browsing | Early, lots of hype | Fix basics; skip gimmicks |
The debate over names: GEO, AEO and SEO
Ahrefs also tracked how “generative engine optimization” searches grew nearly 1,000% over 18 months, while practitioners keep arguing about whether GEO and AEO are new disciplines or SEO with new labels. My view is simple: the label matters less than the work.
Programmatic SEO is back, and so are the penalties
One trend surprised me in Ahrefs’ roundup: interest in programmatic SEO, meaning pages generated at scale from templates and data, is rising again. Ahrefs reports “programmatic SEO” searches grew 124% over 18 months, helped by AI tools that make bulk page creation cheap.
The risk is real. Ahrefs notes that Google’s March and June 2026 spam updates targeted content built for scale rather than users, and it cites Lily Ray tracing more than 70 companies hit in early 2026 updates. Lars Lofgren, also quoted there, warns about companies whose rankings collapsed after scaling to thousands of pages.
Here is how I would think about it:
- Scale useful things, not words. Ahrefs’ Ryan Law argues for scaling free tools, like calculators and checkers, rather than thousands of thin articles.
- Every template page needs unique value: real local data, real prices, real inventory, real reviews.
- Test on a small batch first, and watch indexing and traffic before expanding.
AI makes it easy to publish 5,000 pages in a week. It doesn’t make those pages worth citing. Google’s guidance, quoted above, keeps pointing back to original, people-first content.
Questions business owners ask about AI search trends
Is SEO still worth investing in with AI search growing?
Yes. AI tools still rely on search indexes when they look things up, and Google’s AI features use its normal index. Strong SEO gives AI tools good pages about you to find.
Should I opt my site out of Google’s AI features?
For most businesses, no. Opting out may protect some clicks but removes you from answers your buyers read. Decide based on your business model, and test carefully if you do.
Will ads in ChatGPT replace organic visibility?
Not based on what OpenAI has said. It reports that ads appear below answers and don’t influence them, so being in the answer still matters most.
Do I need an llms.txt file?
It isn’t harmful, but Ahrefs’ study found 97% of llms.txt files are never read. Clear pages and crawl access matter far more.
Which AI search trend should a small business focus on first?
Brand mentions and reviews. They help with AI recommendations, local search and plain word of mouth at the same time.
What to prioritize in AI search next quarter
If I had to pick five things for the next 90 days, they would be these:
- Turn on AI tracking you already have: GA4’s AI Assistant channel and Search Console’s AI reporting. My guide to tracking AI traffic in GA4 shows how.
- Fix your own facts: pricing, specs, policies and comparisons in clear text on your site.
- Earn mentions on the review sites, lists, podcasts and videos your buyers trust.
- Track 50–100 buying prompts and review them weekly with a short action list.
- Run one measurement test, such as a holdout on a group of pages, so you can show what changed.
If your team wants to speed up the data work, my guide to Claude Code for SEO shows how to automate reports safely.
If you want a partner to turn these trends into a plan for your business, our SEO team can help. Get in touch for a quote, and we will start with where your brand stands in AI answers today.
Further reading on AI search trends
- AI Search Strategy: The Four Pillars to Plan Around
- Retrieval-Augmented Generation (RAG) Explained for Marketers
- Ranking Yourself #1 in Your Own 'Best' Lists: Does It Work in AI Search?
Sources for this AI search trends roundup
- Ahrefs – 5 AI Search Trends I’m Seeing in 2026
- Ahrefs – How ChatGPT Does Query Fan-Out (15 Million Queries Studied)
- Ahrefs – AI Overviews reduce clicks (2026 update)
- Ahrefs – How to Measure AI Search Visibility When Attribution Falls Short
- TechCrunch – ChatGPT reaches 900M weekly active users
- PPC Land – OpenAI finally pulls trigger on ChatGPT ads
- Search Engine Land – AI recommendation lists repeat less than 1% of the time
- Google Analytics Help – Default channel group
- Google Search Central – AI features and your website
- Google Search Central – Guidance on generative AI features in Search
