Google AI Mode vs AI Overviews: What Changes for Publishers

Two features, one index

Google began rolling out AI Overviews to everyone in the US on May 14, 2024. They appear inside ordinary search results: a generated summary above the links, with a handful of supporting sources.

AI Mode is a different experience. Google announced on May 20, 2025 that "AI Mode will start to roll out for everyone in the U.S. today." It is a conversational search for questions that need more than a summary.

Google's documentation for site owners draws the distinction this way. AI Overviews help people "get to the gist of a complicated topic or question more quickly" and serve as "a jumping off point to explore links to learn more." AI Mode is for queries that need "further exploration, reasoning, or complex comparisons."

Both are built on Google Search. That is the most important fact in this post.

What Google says it takes to appear

Google's guidance is direct: "There are no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary." A page must be indexed and "eligible to be shown in Google Search with a snippet."

It is equally direct about files and markup: "You don't need to create new machine readable files, AI text files, or markup to appear in these features. There's also no special schema.org structured data that you need to add."

That rules out a whole category of paid advice. Anyone selling AI Mode markup, or a special file for AI Overviews, is selling something Google says does nothing for these features. What matters is what always mattered: pages Google can crawl, index and trust, with content that answers the question.

Query fan-out, and why coverage beats a single keyword

The real difference for publishers is how answers are assembled. Both features may use a "query fan-out technique", which Google describes as issuing multiple related searches across subtopics and data sources to develop a response. Its I/O announcement described AI Mode's version as "breaking down your question into subtopics and issuing a multitude of queries simultaneously on your behalf."

The consequence is that one question can draw on pages that rank for its neighbours, not only the page that ranks for the original phrasing. Semrush found People Also Ask alongside 90.03% of AI Overviews, which fits the same pattern. Google is answering the cluster, not the keyword.

For a content team, that moves the unit of planning from one page per keyword to coverage of a question and everything a person asks next. A pillar page with its cluster is the natural shape for that.

What the click data says

The cost of AI Overviews is well measured. Ahrefs, using December 2025 data on 300,000 keywords, found an AI Overview's presence correlates with a 58% lower clickthrough rate for the top-ranking page. Authoritas found clickthrough fell by close to half on UK news searches, a study Google called "inaccurate and based on flawed assumptions."

How often they appear depends on who is counting. Semrush's tracking of more than 10 million keywords found AI Overviews on 6.49% of queries in January 2025, 24.61% in July and 15.69% in November. BrightEdge's nine-industry tracker found them on nearly half of tracked queries by February 2026.

The shift that matters most for commercial sites is which queries trigger them. In Semrush's data, informational queries made up roughly 91.3% of AI Overview triggers early in 2025 and 57.1% by October, while commercial queries rose from 8.15% to 18.57% and transactional queries from 1.98% to 13.94%. AI Overviews are no longer only a top-of-funnel concern.

AI Mode is harder to measure from outside, because it is a separate experience rather than a panel on the results page. Treat claims about its traffic effects with more caution than the AI Overviews data.

What changes for publishers

1. Expect fewer clicks on answered questions. For informational pages that trigger AI Overviews, plan for impressions to hold while clicks fall, and judge those pages partly on whether they are cited.
2. Plan around question clusters. Fan-out means the neighbouring questions matter. Cover the obvious follow-ups on the page or in linked pages.
3. Answer early, specifically, with sources. The passage a system lifts is usually the one that answers outright. How to Get Cited covers the evidence behind sourced statistics and quotations.
4. Skip the special markup. Google says none is needed. Keep structured data for what it does elsewhere and spend the time on content.
5. Use snippet controls deliberately. nosnippet, data-nosnippet, max-snippet and noindex limit how content appears in AI features, and they are the same controls that govern ordinary snippets, so they carry a cost there too.
6. Watch commercial pages. With AI Overviews spreading to commercial and transactional queries, comparison and pricing pages deserve the attention guides already get.

How to measure it

Search Console shows clicks and impressions for each page. Pages where impressions hold while clicks fall are the first sign an AI Overview is answering for you. That tells you about traffic, not about whether you were cited.

For citations, track the questions directly. AI visibility tracking in Veritas records whether your pages are cited in Google AI Overviews, alongside ChatGPT, Perplexity and Gemini. It does not track AI Mode, and not every tool does, so check before you buy; our AI visibility tools comparison lists what each covers. For the traffic side, see AI Overviews Killed Your CTR.