AI Mode in Google Search: What Happened and How to Track It

AI Mode reached India in July 2025. Here is how to track it in Search Console and GA4, and which of my 2025 predictions came true.

Updated · 9 min read

1234

Key takeaways

  1. AI Mode reached everyone in India in July 2025 and now works in Hindi and seven other Indian languages.
  2. Search Console counts AI Mode clicks in the Performance report. The Generative AI report shows impressions only.
  3. GA4 counts AI Mode visits as Organic Search, so judge them by key events. The work is still SEO.

AI Mode is Google’s chat-style search. It reached everyone in India in July 2025, and it now works in Hindi and seven other Indian languages. To track AI Mode in GSC (Google Search Console), open the Performance report. Its clicks and impressions are counted there under the Web search type, mixed in with normal results.

For a separate view, the Generative AI performance report shows how often your pages appeared in AI Mode and AI Overviews. It counts impressions only. In GA4, AI Mode visits land in the Organic Search channel.

I first wrote this article in May 2025 as ten predictions for Indian SEOs. AI Mode has now been live in India for more than a year. So I have replaced the guesses with what Google actually launched and documented, and checked each prediction against it.

What AI Mode is, and how it reached India

AI Mode answers a question with a written response and links, and you can keep asking follow-ups. Behind the scenes, it may use “query fan-out”. Google’s AI features documentation describes this as “issuing multiple related searches across subtopics and data sources” to build one response.

Here is how it rolled out, from Google’s own announcements:

In its 2025 Year in Search for India, Google says India was the first country outside the US where it launched AI Mode. In an October 2025 post, it calls India its largest user base for voice and visual search.

The old version of this article expected AI Mode in India “within the next few months”. It came faster. The Labs test started four weeks after that post, and the rollout to everyone began six weeks after it.

How to track AI Mode in Search Console and GA4

There is no single AI Mode report. AI Mode data shows up in three places, and none of them isolates AI Mode on its own. Here is what each one shows, and how to read them together.

1. The Performance report: AI Mode clicks, mixed in

Open Performance, then Search results, with the Web search type. Since June 16, 2025, Google has been counting AI Mode in these totals. Its help page on impressions, position and clicks sets the rules:

  • A click on a link from AI Mode to your site counts as a click.
  • Impressions follow the normal rules.
  • Position follows the same method as a normal results page.
  • A follow-up question counts as a new query, with its own impressions, clicks and positions.

What you cannot do is filter AI Mode out. The Search appearance values include nothing for AI Mode, so its clicks sit inside your normal Web numbers. Search Console also leaves out data from Search Labs experiments.

2. The Generative AI performance report: impressions only

This is the closest thing to an AI Mode report. According to Google’s help page, it counts how many times links to your site were shown in AI Overviews and AI Mode. You can break it down by page, country, date and device, and export it.

Know its limits before you put it in a client report:

  • It shows impressions, not clicks.
  • It has no query data, so you see which pages appeared but not for which searches.
  • It does not separate AI Mode from AI Overviews.
  • It may not appear at all if your site has had too few AI impressions.

Google announced the report in June 2026 and says it reached all websites worldwide as of August 31, 2026. A filter splits text searches from multimodal ones, such as Google Lens and Circle to Search. That matters in India: Google says more people use Lens here every month than in any other country.

3. The Search generative AI setting

Under Settings, the Search generative AI control decides whether your site can appear in AI Overviews, AI Mode and generative AI features in Discover. Including your site is the default. If someone switches it to exclude, Google says you get no traffic or impressions from these features. Google also says the setting is not a ranking signal for the rest of Search. Check it once, especially if several people manage your Search Console.

4. GA4: AI Mode visits count as Organic Search

Google’s default channel definitions put AI Overviews and AI Mode in Organic Search. The AI Assistant channel covers tools such as ChatGPT, Gemini and Copilot, and it explicitly excludes them. Google’s documentation does not describe a way to separate AI Mode visits from other Google organic visits. So judge them like the rest of organic search: by key events such as form submissions and calls.

Reading them together

Because no report isolates AI Mode, compare them over the same dates:

  1. Export the pages from the Generative AI report and sort them by impressions.
  2. Look up the same pages in the Performance report. A page gaining AI impressions while its clicks stay flat may be getting answered without a visit.
  3. Split branded from non-branded queries with the branded queries filter. Google added it in November 2025 for top-level properties with enough search data.
  4. Watch long, question-style searches with a Custom (regex) query filter. Google says the average AI Mode search is triple the length of a traditional one. A filter such as ^(\S+\s+){5,}\S+$ shows queries of six words or more. Treat it as a pointer, not as AI Mode data.

The May 2025 predictions, checked

The first version of this article made ten predictions. Two cannot be checked against any public source. One said calculators and other tools would set sites apart; the other said the SEO role would shift to strategy. I have dropped both. Here are the other eight.

Prediction (May 2025)What happenedVerdict
Traditional SEO KPIs become obsoleteGoogle counts AI Mode clicks, impressions and position in the same Performance report, and added an impressions-only Generative AI report.Wrong. The KPIs stayed. They now need context.
Ranking alone won’t be enoughOn an education client in Delhi NCR, my team saw rankings hold for a year while impressions rose and clicks fell (my Adtric post).Right.
Branded search becomes more valuableGoogle added a branded queries filter to Search Console in November 2025. Preferred Sources began rolling out to AI Mode in May 2026.Right direction, and now measurable.
Google properties come firstGoogle’s AI guide says Business Profiles and Merchant Center can help products and services show in AI responses. It also says AI features are “rooted in our core Search ranking and quality systems”.Half right. They help, but your website is still the base.
Content must be built to answer, not just rankGoogle says: “You don’t need to write in a specific way just for generative AI search.” It also says there is no need to break content into small chunks.Overstated. Clear answers help readers. No special format is needed.
Sites must be “AI-ready”, not just user-friendlyGoogle: “There are no additional requirements to appear in AI Overviews or AI Mode”. A page must be indexed and eligible to show with a snippet.Wrong as framed. The bar is the same as for normal Search.
Local and regional-language search gains weightAI Mode added Hindi in September 2025 and seven more Indian languages from October 2025.Right on language. A local boost can’t be measured from public data.
Video and social content shape AI answersGoogle’s AI guide says these features can bring in images and video, and show what is said “including in blogs, videos, and forum discussions”.Partly right. The old claim that Google trains AI Mode on these platforms had no source.

Two pieces of advice in the old version did not hold up. It recommended FAQ and HowTo schema for AI visibility. Google’s guide now says “there’s no special schema.org markup you need to add”, and that its Search ignores llms.txt files. It also suggested special tools to test “AI crawlability”. For Google, normal crawling and indexing is the test.

What I have seen in client work

Search Console does not separate AI Mode from AI Overviews, so nobody can show you a clean AI Mode click count from it. What my team at Adtric can see is the combined effect of Google’s AI answers on client accounts.

In a July 2026 post on the Adtric blog, I described an education client in Delhi NCR. Its rankings had held steady for a year. Impressions kept climbing, and clicks kept falling, quarter after quarter. The SEO had not gotten worse. As I put it then: “The click was simply being answered before it could happen.”

The opposite worry is just as common. A BFSI client of ours wanted to stop SEO after a quarter of reading that AI was killing Google. An AI Overview showed on roughly half of his priority queries. But his brand was cited in it on the queries that mattered most. Branded search was up year on year, and organic conversions were up. Organic also had the lowest cost per qualified lead of any channel he ran. He kept the retainer.

Both cases are about Google’s AI answers in general, mostly AI Overviews, not AI Mode alone. Two habits help here. We now track, for each priority query, whether an AI Overview shows and whether the client is the cited source. In my words on the Adtric blog: “Rank without citation is a half-truth in 2026.” And every month, my team asks real buyer questions in ChatGPT, Gemini and Google’s AI surfaces. We record whether the brand appears.

Google search vs AI Mode in India: what to do now

AI Mode is not a separate search engine that needs a separate plan. It draws on Google’s index and ranking systems, and Google’s guide calls optimizing for it “still SEO”. I agree, with one addition: it is less forgiving of gaps. For ChatGPT, Perplexity and the rest, see my guide on how to rank in AI search.

For Google’s AI Mode, these are the things I would focus on:

  • Be retrievable. A page must be indexed and eligible for a snippet to appear as a link in AI Mode. Fix indexing problems first. My guide to how Google crawls, indexes and decides priority walks through it.
  • Answer the questions around the question. Fan-out turns one query into many. In an Adtric post, I used the example of a buyer asking for the best ready-to-move 3 BHK under ₹2 crore in Noida Extension. The system may look up projects, prices, possession status and locality reviews separately. A site that covers each of those gives Google more to pull from.
  • Write for how India asks. AI Mode questions are longer than classic searches. In India, many are now spoken, and many mix Hindi and English. If a lot of your customers search in Hindi or Tamil, check whether your site gives Google anything in that language.
  • Keep your Google listings current. Google says Business Profiles and Merchant Center can help your products and services appear in AI responses. Keep hours, services and prices up to date.
  • Put a name and real detail on pages. Commodity answers get summarized. First-hand data, named authors and real cases give an AI answer something to cite. My piece on AI-resistant content covers how.

Where to start this month

  1. In Search Console, check that Settings > Search generative AI is set to include your site.
  2. Open the Generative AI performance report and export your top pages by impressions.
  3. Compare those pages’ clicks in the Performance report over the same dates, branded and non-branded.
  4. Fix any important page that Google has not indexed.
  5. List the follow-up questions buyers ask around your main service, and make sure a page answers each one.
  6. Set up GA4 key events on forms and calls, and review Organic Search by leads every month.

If you would rather have someone run this with you, my monthly SEO consulting treats Google and AI search as one plan. It includes a monthly check of whether your site is cited when buyers ask ChatGPT, Gemini and Google’s AI Mode. It is the same SEO work, checked in more places.

Ujjwal Ganesh

Written by

Ujjwal Ganesh

SEO and website consultant with 12+ years of work and 400+ SEO projects, and the author of 25 books for adults and children. I write The Lab from client work and The Living Room from everything else.

Share this article

The method in a book

Websites That Stand Strong

Turn Your Website From a Digital Brochure Into a Growth Engine