AI Mode vs AI Overviews: The Difference and How to Show Up

AI Overviews are AI-written summaries that appear above Google's normal results for some searches. AI Mode is a separate tab that answers in a chat, runs several searches at once, and takes follow-up questions. Both cite web pages; they pick them differently.

Key numbers (as of August 2026)

• AI Overviews reached over 2 billion monthly users by mid-2025, according to Alphabet's Q2 2025 earnings call; Google said at I/O 2025 they were available in 200+ countries and 40+ languages (Google, May 2025).
• AI Mode opened to all US users on 20 May 2025 and passed 1 billion monthly users a year later; Google says AI Mode queries "more than doubled every quarter since launch" (Google, May 2026).
• AI Overviews appeared on 6.49% of keywords in January 2025, 24.61% in July, and 15.69% in November across Semrush's 10-million-keyword sample; the informational share of triggering queries fell from 91.3% to 57.1% over the same period as commercial searches gained (Semrush, December 2025).
• Position one loses ~58% of its clicks when an AI Overview is present (position two −50.8%, position ten −19.4%), across 300,000 keywords of Search Console data, up from a 34.5% drop measured in April 2025 (Ahrefs, February 2026).
• Users who saw an AI summary clicked a traditional result on 8% of visits, versus 15% without one, and clicked a link inside the summary on just 1% of visits (Pew Research Center, July 2025).
• 92.36% of AI Overviews link to at least one domain that ranks in the organic top ten (SE Ranking).
• Google's own guidance: its generative AI features "are rooted in our core Search ranking and quality systems", and "you don't need to create new machine readable files, AI text files, markup, or Markdown" to appear in them (Google Search Central, updated July 2026).

What AI Overviews are

AI Overviews launched in the US at Google I/O on 14 May 2024 as the successor to the Search Generative Experience (SGE) experiment. When Google decides a query benefits from a summary, it generates a few paragraphs above the regular results, with links to the pages the summary drew on. The classic ten blue links, ads, People Also Ask and the rest of the page still render underneath.

Three things about them matter for a brand:

1. They are selective. Semrush's tracking put the trigger rate between 6% and 25% of keywords through 2025, settling around 16% — and the mix shifted from almost entirely informational questions toward commercial and even navigational searches.
2. They draw heavily from pages that already rank. SE Ranking found 92.36% of Overviews link to at least one top-ten domain, and Google says the retrieval uses "our core Search ranking systems".
3. They keep the click. Ahrefs' 58% figure and Pew's 8%-vs-15% figure describe the same effect from two angles: when the answer is on the results page, fewer people leave it.

What AI Mode is

AI Mode is the tab next to "All" in Google Search (and the AI Mode button in the Google app). It launched to all US users on 20 May 2025 after a Search Labs period, ran on a custom version of Gemini 2.5, and by May 2026 was running Gemini 3.5 Flash by default globally.

The mechanism Google describes is query fan-out: AI Mode works by "breaking down your question into subtopics and issuing a multitude of queries simultaneously on your behalf", then synthesising one answer with citations and inviting a follow-up. Deep Search, inside AI Mode, takes the same idea further and can issue hundreds of searches for a single report.

What that means in practice: one AI Mode question might become eight to twelve sub-searches — "best X", "X pricing", "X vs Y", "X for small teams", "X reviews" — and your page only needs to be a strong answer to one of them to be pulled in. It also means AI Mode reaches further down the results than an Overview does, because each sub-query has its own top ten.

AI Mode vs AI Overviews, side by side

How each one picks sources

Google's guide is unusually direct about this. Both features use retrieval-augmented generation: "our core Search ranking systems to retrieve relevant, up-to-date web pages", shown with "prominent, clickable links". AI Mode adds query fan-out, "a set of concurrent, related queries generated by the model to request more information".

So the selection is two-stage:

1. Retrieval — which pages rank for the query (Overviews) or for each generated sub-query (AI Mode). This is ordinary Google ranking, which is why 92% of Overviews cite a top-ten domain.
2. Synthesis — which of those retrieved passages the model actually quotes. Here the content itself decides: a passage that answers the sub-question in one or two clean sentences is easier to lift than one buried in a 400-word preamble.

Ahrefs' study of 75,000 brands adds the brand-level view: across ChatGPT, AI Mode and AI Overviews, the factor that correlates most strongly with being named is branded web mentions (0.664 for ChatGPT, 0.709 for AI Mode, 0.656 for AI Overviews), with YouTube mentions slightly higher still (~0.71–0.74) and backlinks "very weak" (Ahrefs, December 2025). Ranking gets a page retrieved; being talked about on other sites gets a brand named.

How to track whether you appear

• Search Console. Google's July 2026 guide points to a Generative AI performance report in Search Console for AI features. Historically, AI Overview impressions and clicks have been folded into the Web search type, so check what your property actually shows before relying on it.
• Manual checks. Run your 10–20 buyer queries in a signed-out browser and in AI Mode, weekly, and log whether you are cited and which pages are. Answers vary between runs, so sample more than once.
• Trackers. Rank trackers with AI Overview and AI Mode detection, and AI-visibility tools that record citations per query; Sightivo tracks ChatGPT, Claude and Google AI Overviews for your prompt set and shows which sources were cited when you weren't.

How to show up in AI Overviews

Because Overviews cite what ranks, the work is mostly familiar:

1. Rank for the question, not just the keyword. Semrush found triggering keywords skew long and specific; nearly 60% get 100 or fewer searches a month. Cover the question-shaped variants of your topic.
2. Answer first. Put a direct one- or two-sentence answer under each question-shaped H2, then expand. That is the passage the model lifts.
3. Keep the page current and visibly dated. Overviews prefer fresh sources for anything time-sensitive.
4. Structured data where it is honest. Google says it is not required for generative AI features — but it removes ambiguity about which text is the question and which is the answer.

How to show up in AI Mode

Everything above applies, plus two things that follow from fan-out:

1. Cover the sub-questions. Write for "X pricing", "X vs Y", "X for [audience]", "is X worth it" — the queries the model generates — rather than one monolithic page. Each sub-query has its own top ten to be retrieved from.
2. Be mentioned on the pages that get retrieved. AI Mode names brands it has seen across many sources. The Ahrefs correlations are strongest for AI Mode of the three surfaces (0.709 for web mentions). Listicles, comparison posts, Reddit threads and YouTube reviews in your category are the retrieval pool; get into them. That is the part of the job Sightivo is built for: it finds the sources AI engines already cite for your prompts and works out who to contact to get you added.

Google's own summary is the best one-liner: you don't need "to write in a specific way just for generative AI search". You need pages that answer specific questions, and a brand that shows up on other people's pages.

A worked example: one question, two surfaces

Take the query "best project management tool for a 10-person startup".

On the All tab, Google may show an AI Overview: three or four sentences naming a few tools, with two to five link cards drawn mostly from pages already ranking in the top ten for that query — a "best project management software" listicle, a review site, perhaps a Reddit thread. Below it, the usual results. The brands named are the ones those top pages agree on.

In AI Mode, the same question fans out. Behind the answer, Google runs sub-searches along the lines of "project management tools for small teams", "project management software pricing", "Asana vs ClickUp for startups", "free project management tool" — each with its own set of retrieved pages. The answer is longer, names more tools, cites more sources, and invites "which of these has a free tier?" as a follow-up, which triggers another fan-out. A page that never ranks for the head query can still be cited because it is the best answer to one sub-query.

The practical consequence for a startup that is not yet in either answer: for the Overview, you need to be on the listicle that ranks; for AI Mode, you need to be on several of them, and to have pages that answer the narrower sub-questions yourself.

Where Sightivo fits

Sightivo tracks your prompt set across ChatGPT, Claude and Google AI Overviews, shows which competitors and which sources were cited when you weren't, and turns that into a prioritised list of pages to be mentioned on — with the authors behind them and drafted outreach. The free AI visibility checker is the five-prompt version of that first step.

Common Questions

What is Google AI Mode?

AI Mode is a tab in Google Search that answers questions in a conversational format instead of a list of links. It breaks the question into several sub-searches (query fan-out), synthesises one answer with citations, and takes follow-ups. It launched to all US users on 20 May 2025 and passed one billion monthly users by May 2026.

Is AI Mode replacing AI Overviews?

Not as of August 2026. Both exist: AI Overviews appear automatically above normal results on a subset of searches; AI Mode is a separate tab the user chooses. Google has said users will "continue to get a range of results from Search". Features do migrate from AI Mode into Overviews over time.

How do I turn AI Mode on or off?

You open it by tapping the AI Mode tab in Google Search or the button in the Google app. There is no setting that turns AI Mode off for you as a searcher; searching from the "All" tab gives you the classic results (possibly with an AI Overview).

Does AI Mode use the same sources as AI Overviews?

Largely the same index, selected differently. Both retrieve from Google's Search index using its core ranking systems. AI Overviews retrieve for the one query, so they lean on the organic top ten (92.36% cite at least one top-ten domain, per SE Ranking). AI Mode fans the question out into several sub-queries, each with its own top results, so it reaches a wider set of pages.

How do I rank in AI Mode?

Rank for the sub-questions AI Mode is likely to generate — pricing, comparisons, audience-specific and "is it worth it" queries — with a direct answer at the top of each section, and get your brand mentioned on the listicles, review sites, Reddit threads and YouTube videos that those sub-queries retrieve. Branded web mentions correlate more strongly with AI Mode visibility (0.709) than any other factor Ahrefs measured except YouTube mentions.

Does AI Mode reduce clicks?

AI Overviews demonstrably do: Ahrefs measured a ~58% drop in position-one CTR and Pew measured 8% vs 15% click rates with and without a summary. AI Mode has not been measured at the same scale yet; as a whole-page answer with follow-ups, it is reasonable to expect fewer and later clicks, concentrated on the pages it cites. Google reports its AI Mode queries doubled every quarter since launch but has published no click data.

Key Takeaways

• AI Overviews summarise above the results; AI Mode replaces the results with a chat that fans your question out into many searches.
• Both retrieve from Google's normal index — 92% of Overviews cite a top-ten domain — so ranking still gets you into the pool.
• Being named is a different signal: branded web mentions correlate ~0.66–0.71 with visibility across Google's AI surfaces; backlinks barely register.
• Clicks fall when an answer is on the page: −58% for position one under an Overview, 8% vs 15% in Pew's panel.
• Write answer-first sections for the sub-questions, keep pages dated and current, and get mentioned on the sources these surfaces already cite.

See where you stand today with the free AI visibility checker, or read the full generative engine optimization guide for the complete playbook.