Google AI Mode E-Commerce Expansion: What to Do

What Google actually changed with AI Mode

AI Mode is the conversational, Gemini-powered tier of Google Search that launched publicly in the US on May 20, 2025, then expanded to 180 countries and added agentic shopping capabilities through late 2025 and early 2026.

People are coming to Google to ask more of their questions, including more complex, longer and multimodal questions. Google - blog.google

Practically, AI Mode is a separate tab at the top of Google Search (positioned to the left of the standard 'All' results tab) that handles longer, multi-part, conversational queries. Under the hood it uses 'query fan-out' - breaking one question into subtopics, running parallel searches, and synthesizing one synthesized answer with a sidebar of source domains. For shopping queries, AI Mode pulls from Google's Shopping Graph (50 billion product listings refreshed at 2 billion per hour, per Lilian Rincon's I/O 2025 announcement) and renders a panel of product listings alongside the synthesized response.

The data from independent studies is consistent. Ahrefs documented that AI Overviews reduced clicks by 34.5% as of April 2025 and that AI Overviews appear on 54.61% of all US Google searches; AI Mode behavior compounds this further. Kevin Indig's analysis of 250 search tasks found a near-100% zero-click rate within AI Mode itself, meaning users do their research-and-comparison work entirely inside the AI interface and only click through when they are ready to transact. For ecommerce, that shifts the click distribution toward late-funnel intent and pulls top- and mid-funnel engagement off-site.

When the rollout took effect

AI Mode launched publicly in the US (English) on May 20, 2025. International expansion to 180 countries was announced through Q3 2025. The shopping experience with agentic checkout (Ticketmaster, StubHub, Resy, Vagaro as launch partners) rolled into AI Mode in the US during late 2025. AI Overviews coverage separately reached over 200 countries and 40 languages by the same period.

• Announced: May 20, 2025 at Google I/O (Liz Reid, VP and Head of Google Search)
• Rollout begins: May 20, 2025 (US English) for all US Google Search users
• Full rollout: 180 countries (announced Q3 2025), shopping features through Q1 2026
• Enforcement begins: ongoing - international and language coverage expanding throughout 2026

Who this actually hits

The most exposed segment is ecommerce sites whose traffic concentrates in informational, research, and comparison queries. Branded and direct-purchase intent traffic is much less affected because AI Mode currently reserves clicks for the transactional moment. The pattern is consistent with independent research from Ahrefs, Kevin Indig, and Adobe.

If your traffic concentrates in branded queries or direct-product searches you are largely insulated, and what traffic you do receive from AI surfaces tends to convert better. Adobe reported AI-referred traffic to US retail sites grew 4,700% year-over-year and that AI-referred visitors spend 32% longer on site than visitors from paid, email, organic, or social. For YMYL ecommerce (health supplements, financial products, regulated goods), the E-E-A-T evidence requirements are higher and AI Mode citation tends to favor expert-bylined content over pure transactional pages.

What to do this week

Priority order: segment by query intent, identify the cohort losing the most clicks, audit your evidence surface, and only then start rewriting pages. AI Mode visibility is not reportable in Search Console (Google lumps AI Mode traffic into the standard Search bucket), so the diagnosis has to come from query-intent segmentation, not from GSC's AI-feature filter (it does not exist).

1. Segment GSC traffic by query intent. Export Search Console performance data for the last 12 months. Tag queries as branded, transactional, comparison, informational, or how-to. Compare click trends by tag, not by URL. The decline lives in comparison and informational queries; branded and transactional should be stable or up. Per Marie Haynes' diagnostic framework, segment the data before you change anything.
2. Audit where AI Mode is citing competitors instead of you. For your top 20 commercial queries, run AI Mode searches yourself and screenshot the cited domains. Per the Semrush AI Mode study, ~89% of cited link domains overlap with Google's top 10 organic for the same query, but only ~51% of cited URLs match - meaning AI Mode picks different pages from the same domains. If you are ranking but not cited, the question is page-level, not site-level.
3. Strengthen your Merchant Center feed and product schema. The Shopping Graph (50 billion product listings refreshed hourly) is what populates the AI Mode shopping panel. Audit your Google Merchant Center feed for GTIN, brand, condition, availability, and GS1 Digital Link compliance. Cross-check Product schema on the PDP for completeness (offers, aggregateRating, image array). Our breakdown is in product schema optimization.
4. Build third-party citation surface. Per Aleyda Solis's five-subvertical AI search study, AI engines cite guides, support content, reviews, communities, marketplaces, videos, and expert media alongside transactional pages. Audit your brand mentions on Reddit, YouTube, and review platforms. If AI Mode is summarizing those surfaces instead of your PDPs, that is where your investment goes next. Methodology in how Reddit, YouTube, and Wikipedia dominate AI citations.
5. Track AI visibility with a dedicated tool. Because Google does not expose AI Mode performance in Search Console, you need a separate visibility stack. Ahrefs Brand Radar, Semrush Enterprise AIO, Profound, and Similarweb AI Traffic all approach this differently; pick one and baseline now so you can measure delta when the next AI Mode feature ships. Our setup guide is in how to use Ahrefs Brand Radar to monitor AI visibility.

What to do this quarter

The strategic shift, in one line: AI Mode and AI Overviews consume top- and mid-funnel engagement, leaving the bottom-of-funnel transactional click. Per Glenn Gabe's kitchen-sink framing on Google updates, the response is not single-lever - it spans content depth, knowledge graph, evidence surface, and feed quality at the same time.

Move beyond PDPs and PLPs as the only optimization target

Per Aleyda Solis's analysis of five US ecommerce subverticals, AI engines often cite the pages that help resolve buyer uncertainty before, around, and after the purchase - guides, size and fit resources, comparison pages, FAQs, review content, and third-party media. The pages users actually visit from AI platforms are not always the same pages AI systems cite as evidence. Plan for both: cited evidence pages and clickable transactional pages. The full breakdown of this distinction lives in our piece on GEO for DTC brands.

Build a brand knowledge graph entity

Per Ahrefs' AI Overviews study, brand web mentions show the strongest correlation (0.664) with AI Overview brand visibility, and brands in the top 25% for web mentions earn over 10x more AI Overview citations than the next quartile. Building an Organization knowledge graph (Wikipedia, Wikidata, Crunchbase, LinkedIn Company, sameAs schema) and earning category mentions in third-party media moves the needle on AI Mode citation eligibility. Our guide is at how to create a brand knowledge graph that LLMs can understand.

What we're seeing in real accounts

Note: the patterns below are aggregated and anonymized from audits where we have diagnosed AI Mode and AI Overviews impact. In the accounts we have reviewed, the split between query types holds consistently: branded and product-name traffic tends to stay flat or grow, while category-level informational traffic ("best X for Y", "how to choose X") often drops in the 25% to 45% range over the 12 months following the AI Mode US launch.

Counterexample: B2B SaaS accounts whose traffic concentrates in branded plus category-defining searches ("[brand] alternatives", "[brand] pricing") have been largely insulated. Ahrefs published a similar pattern internally - AI search drives less than 1% of total traffic but is their highest-converting channel at 10%+ conversion rate. The lesson is not "abandon top of funnel" but "recognize that the surfaces you optimize for top of funnel are now AI surfaces, not search result pages."

What we're still watching

AI Mode and the AI Overviews tier are still in active rollout. Four open questions are driving how we sequence audit work for the next two quarters.

• AI Mode reporting in Search Console: Whether Google adds AI Mode performance as a separate filter in Search Console. Currently AI Mode traffic is lumped with regular search per Barry Schwartz, which makes attribution analysis impossible. A dedicated filter would unblock honest client reporting.
• Ad expansion in AI Mode: SEO consultant Brodie Clark documented ads now appearing inside AI Mode (cited in Semrush's AI Mode roundup). Whether ad density inside AI Mode increases the way it did in traditional SERPs will reshape the click-to-AI-citation balance again.
• Shopping Graph optimization signals: Whether Google publishes more guidance on how merchants can rank inside the AI Mode shopping panel (currently this is opaque - it sits between Merchant Center feed quality and the underlying ranking systems). The next Search Central update on Shopping Graph will likely change the optimization playbook.
• International rollout pace: Whether AI Mode reaches non-English markets at the same intensity. AI Overviews has reached 200+ countries and 40+ languages but AI Mode (the conversational tier) expansion has been slower. International ecommerce sites should baseline AI Mode visibility in their markets now.