Google search is no longer just a list of blue links. By 2026, a large portion of searches will already return a ready-made answer at the top of the page, assembled by artificial intelligence. This is the effect of Google with AI, namely AI Overviews, which summarize the answer above the traditional results, and AI Mode, a conversational search that investigates the topic in several stages and cites the sources it used.
For those who sell online, the consequence is direct. The customer finds what they were looking for without needing to enter your website. And this changes the organic traffic calculation that most stores were getting until recently.
The magnitude of the change is already apparent in the data. A Pew Research Center study , published in July 2025, tracked the actual browsing behavior of 900 users in the United States. When an AI summary appeared in the search results, the click-through rate on links dropped from 15% to 8% of visits. And only 1% of people clicked on the source cited within the summary itself. The information was consumed on the results page, and the visit ended there.
On this page you will find:
Google with AI: what really matters
- Google with AI brings together two formats: AI Overviews (summaries generated at the top of the results) and AI Mode (Conversational search that responds in stages and links to sources).
- Organic traffic tends to drop when an AI summary appears: in the study of Pew Research CenterClicks on links fell from 15% to 8% of visits.
- There is no separate AI index: Generative responses use the same index as traditional search, so anyone who isn't trackable on Google isn't mentioned by the AI either.
- A page becomes a quotable source when it answers questions clearly and independently., with concrete data and content visible in the HTML.
- Several shortcuts don't work: Google ignores the llms.txt file, it doesn't recognize any "schema for AI"and treats phrases designed to please robots as spam."
- The foundation remains well-done technical SEO.Useful content and structured data consistent with what the user sees on the screen.
What changes in organic e-commerce traffic?
The first point to understand is that fewer clicks doesn't mean less importance ; it means a different kind of competition. Before, the goal was to rank in the top positions to receive clicks. Now, in addition to that, your store needs to be the source that the AI chooses to assemble the response.
When Google's AI summarizes a topic and cites three or four pages, those brands gain visibility and authority even without an immediate click. Those not cited disappear from the conversation. Therefore, the benchmark for success is no longer just ranking position, but also includes presence in the generated results.
The second point is the behavior of those searching for products . Informative searches (“ what is the difference between fabric A and B ”, “ how does free shipping work ”) are the most absorbed by AI summaries. However, searches with a clear purchase intent, including product name and brand, still lead the person to the store. This reinforces a layered content strategy: educating at the top and capturing purchase intent in the middle and bottom of the funnel.
How Google uses AI to decide which pages to cite.
Google has published, for the first time in an organized manner, a guide on what websites need to do to appear in these formats . The recommendations confirm something important: there's no new magic, just discipline.
Technical SEO is a prerequisite.
If Googlebot cannot access the page, it will not appear in any search results, whether generated or not. In practice, this means no accidental blocks in robots.txt, no server errors, no noindex tag, and the content must be present in the rendered HTML, not hidden behind a click or JavaScript that only loads later. Core Web Vitals, mobile version, and HTTPS remain valid requirements.
Content in response format
AI summaries are assembled from snippets that Google extracts from indexed pages, following the same logic as featured snippets. Titles that represent real questions, answered in a few sentences below, have a better chance of becoming a source. The sentences need to make sense on their own: expressions like " as explained above " hinder extraction because the system takes the snippet out of context.
Concrete data and consistent structure
Generative systems prefer precise statements to sales language. Numbers, definitions, and update dates help. In e-commerce, structured data (Schema.org) gains special importance for 2026: product tagging with price and availability, local business data, and organizational information. The rule is consistency; if the tagging states one price and the page shows another, Google treats the discrepancy as manipulation.
An additional precaution should be taken for stores with many similar URLs: without a well-defined canonical tag and consistent redirects, Google splits the signals between versions and does not choose any as the preferred source.
➡️ LEARN MORE: What is technical SEO?
Metrics that start to make sense
Tracking only organic sessions and average position is insufficient. It's worth including the organic click-through rate by search type , to separate what still brings visits from what has become zero-click. It's worth measuring the rate of searches that end without a click , since the Pew study itself showed that 26% of sessions with AI summaries ended there, compared to 16% without the summary . And it's worth monitoring in which Google AI-powered answers your brand is mentioned , a new metric that functions as voice share within generative search.
These numbers tell a story that ranking position alone no longer tells: whether your store is present where the decisions are actually made.
If your organic traffic has started fluctuating without a clear explanation, some of the answers are likely happening within Google itself. It's worth mapping where your brand is being left out and adjusting content, technical structure, and structured data to get mentioned again. The e-Plus Media Management and SEO team works with this type of diagnosis daily, and you can find more materials about search and content here on the blog to start looking at your own data.
FAQ
What is Google with AI?
Google with AI is the set of generative search formats in Google, consisting of AI Overviews and AI Mode. AI Overviews summarize the answer at the top of the results. AI Mode is a conversational search that answers in steps and cites the sources used to assemble each answer.
Do AI Overviews reduce my website's organic traffic?
They reduce clicks on informational searches, primarily. According to the Pew Research Center, when an AI summary appears, the click-through rate on links drops from 15% to 8% of visits. Searches with a clear purchase intent still lead the user to the store, which preserves a good portion of transactional traffic.
How can I get my e-commerce site mentioned by Google's AI?
The foundation is to be crawlable and indexable, with content visible in the HTML. From there, structure the text as straightforward questions and answers, use concrete data, and apply structured product and organizational data consistently across the page. There are no shortcuts or exclusive markup for AI.
Do I need to create an llms.txt file to appear in Google using AI?
No. Google has confirmed that it ignores the llms.txt file: it neither helps nor hurts. Similarly, there is no such thing as an "AI schema," and phrases advertising optimization for language models are treated as spam. What works is solid SEO and useful, original content.
