Key takeaways
- AI Mode surpassed 1 billion monthly active users one year after launch, with queries more than doubling every quarter, according to Google's own I/O 2026 reporting.
- Google isn't killing Google Shopping. It's wiring AI Mode, the Gemini app, and even YouTube into the same product feed through a new Universal Cart and the Universal Commerce Protocol (UCP).
- Promptwatch's June 2026 citation data shows google.com alone makes up 7.31% of all AI Mode citations, more than YouTube and Reddit combined, which tells you AI Mode leans hard on Google's own shopping and local inventory before it ever leaves the building.
- Product pages have become the most-cited content type across both ChatGPT and Google AI Overviews in 2026, overtaking listicles for the first time in late July, so brand-owned product pages matter more than ever regardless of which surface wins.
- Treat this as an evolution of Shopping, not a replacement. The underlying Merchant Center feed still powers both the classic Shopping tab and whatever AI Mode decides to surface.
So is AI Mode actually replacing Google Shopping
Short answer: no, not in the sense of one thing switching off and another switching on. Longer answer: Google is doing something more interesting, and arguably more threatening to third-party retailers, which is folding Shopping's entire backend into AI Mode as a new front door.
At Google I/O 2026, Liz Reid (VP, Head of Search) announced that AI Mode had crossed 1 billion monthly active users just a year after launch, with query volume more than doubling every quarter since. That's not a feature rollout anymore, that's a second search engine running inside the first one. And Google used the same keynote to unveil a Universal Cart, a cross-merchant shopping cart that works inside Search, the Gemini app, and soon YouTube and Gmail.
The launch retailer list for Universal Cart reads like a mall directory: Nike, Sephora, Target, Ulta Beauty, Walmart, Wayfair, plus a batch of Shopify merchants including Fenty and Steve Madden. It started in the US through the Gemini app and Search in summer 2026, with YouTube and Gmail integration planned shortly after.
Here's the part that actually matters for the "replacing" question. Google didn't build a parallel shopping engine for AI Mode. It built a new checkout layer, the Universal Commerce Protocol (UCP), that sits on top of the same Merchant Center feed that's always powered the Shopping tab. In November 2025, agentic checkout completed purchases on the merchant's own site using Google Pay. By January 2026, UCP moved the transaction onto Google's own surfaces, with the merchant becoming "merchant of record" instead. By May 2026, that same UCP-powered checkout started showing up on ordinary, non-AI-Mode Shopping listings too. So the infrastructure converged before the surfaces did.
That's the pattern across almost everything Google has shipped in 2026: AI Mode isn't a rival to Shopping, it's a new interface sitting on the same plumbing, plus a conversational layer and an agent that can act on your behalf.
What the citation data actually shows
This is where things get less flattering for the open web, including for the brands that aren't Google itself.
Promptwatch's citation share data for Google AI Mode in June 2026 found that google.com alone accounts for 7.31% of all AI Mode citations, more than YouTube (2.88%) and Reddit (2.52%) combined. Add YouTube back in and Google-owned properties absorb over 10% of the entire citation mix. That means a meaningful share of AI Mode answers get resolved using Google's own Maps listings, Business Profiles, and support docs before the system ever reaches for a third-party page.
This self-preference is specific to AI Mode. In ChatGPT's June 2026 citation data, Reddit leads the pack and google.com barely shows up at all. Different model, very different sourcing behavior.
The month-over-month trend makes the picture clearer, not murkier. google.com's AI Mode citation share climbed from 4.24% in May to 7.31% in June, a 72% jump stacked on top of a sixfold increase before that. Meanwhile, no independent website clears even 1% of AI Mode's citation share. The open web, in other words, is a long tail sitting underneath Google's own properties, not a level playing field.
For anyone selling product through organic visibility, that's the real headline buried under the "AI Mode vs. Shopping" framing. AI Mode isn't just changing where shopping happens, it's changing who gets to answer the question in the first place, and right now the answer is increasingly "Google."
Product pages are winning the content war, regardless of surface
If there's good news for retailers in this data, it's that the content type that wins inside AI answers is squarely in their wheelhouse: product pages.
Promptwatch's content-type breakdown for ChatGPT citations found product pages are the single most-cited content type overall at 22.4% of all classified citations, with landing pages close behind at 16.8%. Together that's 39% of all citations going to pages that describe something a company sells. Narrow it to commercial or comparison prompts specifically and product pages climb to 26.0%.
Google AI Overviews tells a similar story. Promptwatch's AI Overviews citation types report for July 2026 shows product pages overtaking listicles as the most-cited format for the first time ever, starting July 28, 2026, and closing the month at 17.9% versus 16.2% for listicles. Product-page citation share nearly doubled since January 2026, going from roughly 9% to 16-18%, while listicles fell from an average of 26% in Q1 down to 18%. That's a real shift in how AI search treats commercial content: brand-owned product pages are beating third-party "best of" roundups for the first time.
So whether a shopper ends up in AI Mode, AI Overviews, or ChatGPT, the content doing the heavy lifting is increasingly the product page itself, not a blog post about the product.
How ChatGPT's shopping ambitions compare
It's worth looking sideways at what OpenAI is doing, because it puts Google's moves in context. Promptwatch's ChatGPT shopping feature usage data shows shopping features like product cards and price comparisons appear in only a low single-digit percentage of ChatGPT web-search responses overall, concentrated on clearly transactional prompts. The frequency moves in sudden step-changes rather than drifting up gradually, roughly doubling overnight in late May 2026 before dropping back weeks later, which suggests OpenAI is manually tuning when shopping triggers rather than letting it emerge organically.
Promptwatch's own framing on this is worth repeating: treat ChatGPT shopping the way you'd have treated Google Shopping in its earliest days, modest traffic today, but the brands that establish product visibility before the feature scales will be the hardest to displace once it does.
There's a monetization angle here too. ChatGPT showed zero ads until May 27, 2026. Per Promptwatch's ads-over-time data, ad frequency then ramped in steps, a spike in late June, then a sustained jump above 30% of citation-enabled responses starting July 1. By mid-August the 7-day average sat at 32.4%, up 12.5 points week over week. Most of those ad-bearing responses, 73.3%, come from organic commercial prompts rather than brand-name lookups, meaning OpenAI is monetizing broad shopping intent the same way Google has for two decades, just starting from a much smaller base.
Side-by-side: AI Mode shopping vs. classic Google Shopping vs. ChatGPT shopping
| Dimension | Classic Google Shopping | Google AI Mode shopping | ChatGPT shopping |
|---|---|---|---|
| Checkout | Merchant site or Google Pay | Universal Cart + UCP, transaction can complete on Google's own surface | Product cards, no native universal cart yet |
| Data source | Merchant Center feed | Same Merchant Center feed, plus conversational attributes | Retailer feeds plus general web crawl |
| Citation behavior | N/A (structured listings, not citations) | Heavily self-referential: google.com + YouTube over 10% of citations | Low single-digit response share, step-changes in frequency |
| Monetization | Mature Shopping ads business | AI performance insights in Merchant Center, ads still forming | Ads appeared May 2026, now 30%+ of citation-enabled responses |
| Rollout stage | Fully mature | Early access, US-first, expanding to Canada/Australia | Early, feature toggled on and off |
What this means if you sell products online
A few practical implications fall out of all this data, and none of them require you to bet on one surface winning outright.
First, your Merchant Center feed is no longer just powering a shopping tab, it's powering the conversational answer too. Google's new Conversational Attributes schema, previewed at Google Marketing Live 2026, is specifically designed to make product feeds more legible to AI Mode's language model layer. If your feed is thin, that thinness now shows up in two places instead of one.
Second, product pages are the asset to invest in, not blog content about your products. Both ChatGPT and AI Overviews now favor product pages over listicles for commercial queries, a reversal from where things stood in early 2026. If your product pages read like they were built for a 2015-era shopping crawler rather than a model trying to answer "what's the best waterproof hiking boot under $150,” that's the gap to close.
Third, don't assume AI Mode is a neutral discovery layer. With google.com and YouTube together absorbing over 10% of AI Mode's citations, and no independent site clearing 1%, the system has a structural preference for keeping users inside Google's own properties before it reaches for yours. That's worth knowing before you pour budget into "ranking" inside AI Mode as if it behaved like classic organic search.
Fourth, keep an eye on ChatGPT's shopping rollout even though it's small today. Ad frequency jumping from 0% to over 30% of commercial responses in under three months is not the behavior of a feature still in testing, it's the behavior of a feature Open AI intends to scale fast. Establishing product visibility there now, while competition is thin, is cheap insurance.
If you want to track any of this systematically, that's really the job of an AI visibility platform rather than manual spot-checking. Tools like Promptwatch track citation share across AI Mode, AI Overviews, ChatGPT, Perplexity and the rest, including a dedicated ChatGPT Shopping and Ads Radar for watching exactly these kinds of step-changes in real time.
For ecommerce brands specifically, there are also category-focused tools worth a look, such as Alhena AI for AI visibility tracking built around product catalogs, Zoovu for on-site AI product discovery, and Yotpo Discover for catalog and review-driven AI visibility. If you'd rather browse the full landscape of GEO-focused tools before picking one, the directory at bestgeosoftware.com is a reasonable place to start.
The bottom line
AI Mode is not a replacement for Google Shopping, it's an absorption of it. The checkout infrastructure, the feed requirements, and the ad business are converging into one system that happens to have two front ends right now, a classic Shopping tab and a conversational AI Mode layer, with the Gemini app and YouTube joining soon. What the data makes clear is that Google isn't choosing between these surfaces, it's making sure the same inventory and the same monetization model work across all of them.
The bigger risk for third-party retailers isn't that AI Mode kills Shopping. It's that AI Mode, by its own citation data, leans on Google's own properties more than any other AI surface does, which means winning visibility there requires a different playbook than winning a position in the old Shopping tab ever did. Product pages matter more. Structured, conversational-ready feeds matter more. And the window to establish visibility before ChatGPT's own shopping ambitions mature is getting shorter every month.