Google AI Overviews for Ecommerce: What Happens to Your Traffic When AI Answers the Search

Your product page can rank number one and still lose most of its clicks. When Google shows an AI Overview, the top organic result now earns about 58% fewer clicks than it did before the feature existed, according to Ahrefs. The shopper still searched. Google simply answered first.

For ecommerce and Amazon brands, this is not a publisher problem anymore. AI Overviews jumped from 2.1% of shopping queries in November 2025 to 14% by March 2026, per a Visibility Labs study of 20.9 million shopping results. On "best [product]" research searches, the share is far higher.

This guide covers what the data says about lost traffic, which ecommerce searches are most exposed, where Google pulls its answers from, why the clicks that survive are worth more, and a practical playbook for brands selling on their own sites and on Amazon.

What AI Overviews and AI Mode actually do on a shopping search

An AI Overview is a Gemini-written summary that sits above the organic results and cites a handful of source links. Pew Research found the typical summary runs about 67 words, and 88% cite three or more sources. For a shopper, that is often enough to shortlist products without opening a single page.

AI Mode goes further. It is a full conversational search tab that, as of May 2026, had crossed 1 billion monthly users, per a roundup of Alphabet's figures. On product searches it pulls from Google's Shopping Graph, which holds more than 50 billion listings with over 2 billion updated every hour, according to Channable.

Three things now compete for the shopper's attention above your organic listing:

• The AI answer itself, which recommends products and explains trade-offs.

• Shopping units and ads, which keep taking click share from classic organic results.

• Agentic checkout, where Google can track a price and complete the purchase through Google Pay on the merchant's behalf (BigCommerce).

The position-by-position picture matters for ecommerce. Ahrefs measured a click loss on every page-one slot, from about 50.8% at position 2 to 19.4% at position 10. Ranking well no longer guarantees the visit.

The citation itself rarely saves you either. Pew found shoppers clicked a link inside the AI summary in just 1% of visits. Searchers also ended their browsing session on 26% of AI Overview pages, against 16% on standard results.

Across all of Google, SparkToro estimates 68% of US searches ended without a click in early 2026. AI Mode is reported to push the zero-click rate as high as 93% (Lilach Bullock), so expect the baseline to keep rising as adoption grows.

Pew's data explains why. Longer, question-style searches trigger summaries far more often: 53% of searches of 10+ words got one, against 8% of one- or two-word searches. Shoppers comparing "which protein powder is best for lactose intolerance" will see an AI answer. Shoppers typing your brand name mostly will not.

What this means in practice:

• Most exposed: buying guides, "best X for Y" listicles, comparison posts, how-to and FAQ content. This is the top-of-funnel content many DTC brands built their SEO on.

• Moderately exposed: generic category and product pages, as AI Overviews keep expanding into shopping results.

• Least exposed: branded searches and high-intent transactional queries, which SparkToro also found still drive meaningful clicks.

Where AI Overviews get their answers

Google does not just quote the top-ranking product page. It builds answers from a wide mix of sources, and much of that mix is outside your website.

• Community and video platforms. Wikipedia, YouTube and Reddit were the three most cited sources in Pew's study, together making up 15% of AI summary citations.

• Retailers and marketplaces, but less than you'd think. In an LLM Pulse study of 391,000 US shopping citations, Amazon was the most cited retailer, yet appeared in only 5.53% of shopping answers. Walmart followed at 1.97%. Being listed on a big marketplace does not automatically put you in the answer.

• Product data feeds. For shopping results, Google's Shopping Graph draws primarily on Merchant Center feeds, with Schema.org Product markup on your pages used to verify that data (paz.ai).

One more caution: AI answers are unstable. An academic audit of AI product recommendations found Gemini's displayed sources overlapped only 14.8% between identical queries submitted moments apart. Visibility in AI search is a share you win repeatedly, not a ranking you hold.

The upside: fewer clicks, better buyers

The traffic you lose is mostly browsing traffic. The traffic you keep, and the traffic AI sends you, is increasingly ready to buy.

Being cited pays off. Seer Interactive found brands cited in an AI Overview earned 120% more organic clicks per impression than uncited brands on the same results (Seer Interactive). On transactional queries where an Overview appeared and the brand was cited, organic CTR more than doubled across 2025, from 0.7% to 1.7% (Herdl).

Citation lifts your ads too. On informational queries, cited brands averaged a 15.74% paid CTR against 11.19% for uncited brands (Pixis). Organic and paid visibility now reinforce each other.

AI-referred shoppers convert better. In July 2026, AI-referred visits to US retail sites converted 60% better than non-AI traffic, the 11th straight month of outperformance, per Adobe data reported by Digital Commerce 360. That is a full reversal from March 2025, when AI traffic converted 38% worse (Decrypt).

The logic is simple. A shopper who reads an AI comparison and then clicks through has already done their research. They arrive further down the funnel, so each visit is worth more even as total visits shrink.

The playbook: how ecommerce and Amazon brands should adapt

The goal has shifted from ranking under the answer to being named inside it. Here is where to focus, in rough order of impact.

1. Treat your product feed as your most important SEO asset. Keep Google Merchant Center complete and current: GTINs, accurate price and stock, quality images, product category and type. Turn on free Shopping listings, which many paid advertisers leave switched off. In May 2026 Google also added optional "conversational attributes" to Merchant Center, built for AI-style queries (SEOquick).

2. Make your site and feed agree. Add Schema.org Product markup to every product page, pulling from the same source as your feed. When price, availability or ratings disagree, Google can deprioritize both, and mismatches increasingly trigger misrepresentation suspensions.

3. Rebuild research content to be quotable. Answer the question in the first two sentences, then back it up with specifics: dimensions, materials, use cases, honest trade-offs and comparison tables. AI models cite pages that state clear facts, not pages that bury them under fluff.

4. Target the long, specific questions your buyers ask. "Best stainless steel water bottle for kids under 500 ml" is where AI Overviews live. Build pages that answer those exact use cases better than a generic listicle can.

5. Earn mentions where AI looks. Reddit, YouTube and review sites are heavily cited. Genuine participation in relevant subreddits, creator reviews and third-party comparisons all feed the sources AI draws on. Off-site reputation now directly shapes on-SERP visibility.

6. Strengthen your Amazon listings, not just your website. Amazon is the single most cited retailer in AI shopping answers. Clear titles, attribute-complete listings, strong A+ content and a healthy review profile make your product easier for any AI to describe accurately.

7. Protect your branded and transactional terms. These still convert through classic results. Make sure brand searches land on a fast, complete page, and use paid search where competitors bid on your name.

8. Prepare for agentic checkout. Google's agentic checkout lets AI complete purchases while you stay the merchant of record. That requires accurate real-time inventory, clear shipping and returns policies, and a Merchant Center account in good standing.

How to measure the impact on your store

Clicks alone now understate your visibility, so track impressions, citations and revenue side by side.

• Look for the "crocodile mouth" in the Search Console. Rising impressions with flat or falling clicks on a query usually means an AI Overview is answering it. Segment by query type to see whether losses sit in research content or product pages.

• Tag your queries by intent. Split informational, commercial and branded queries. Expect the first group to lose clicks and the last to hold up. If branded clicks fall, investigate that first.

• Track AI citation rate as its own KPI. Report organic CTR and AI citation share separately rather than as one blended "visibility" number. Sample your top 50 revenue queries monthly and record whether you are cited.

• Measure AI referral traffic in analytics. Create a channel group for visits from Gemini, ChatGPT, Perplexity and similar sources. Compare conversion rate and revenue per visit against organic search.

• Judge SEO by revenue, not sessions. A 30% drop in blog traffic with stable product-page revenue is a very different story from a drop in sales.

Frequently asked questions

Do AI Overviews appear on product searches? Increasingly, yes. They appeared on about 14% of shopping queries by March 2026, up from 2.1% four months earlier, and on roughly 83% of "best [product]" queries.

How much traffic will my store lose? It depends on your mix. Studies put the click loss at roughly 40% to 60% on searches that show an AI Overview. Stores relying on buying guides and blog content feel it most; stores driven by branded and transactional searches feel it least.

Can I opt out of AI Overviews? Not without hurting your search visibility. Blocking Google from using your content generally means limiting how it appears in Search, so for most stores the better move is to compete for citations.

Does being on Amazon help me appear in AI Overviews? It helps, since Amazon is the most cited retailer in AI shopping answers. But it is cited in only about 5.5% of answers, so your own site, product feed and off-site mentions still matter.

Is SEO dead for ecommerce? No. It has shifted from chasing rankings to earning citations, clean product data and high-intent visits. Brands doing that well are getting fewer but more valuable clicks.

The bottom line

AI Overviews are not killing ecommerce search. They are filtering it. The casual researcher now gets an answer on Google, while the shopper who clicks through arrives informed, confident and far more likely to buy.

The brands that win will be the ones Google trusts enough to name: clean product data, content that answers real buyer questions, a strong presence on Amazon and in the communities AI reads, and measurement that values revenue over raw sessions.

If you sell into the US from India, the UK or Europe, this shift is a chance to compete on clarity rather than budget. Brandkyte helps Amazon and ecommerce brands get found, cited and chosen as search moves to AI. Get in touch to see where your brand stands.