Over three weeks (21 days) in July 2026, we analyzed more than 2 million product listings across more than 100,000 SERPs and AI Mode responses, running the same query on both surfaces on the same day. On identical queries, AI Mode returned roughly 95% fewer product listings than standard search.
Google is rolling AI Mode out quickly and building it into its main products at a rapid pace. If AI Mode becomes the default in its current form, most of the Google Shopping visibility that stores worked hard to earn simply disappears. And this is not some far-off scenario: AI Mode already opens right from the normal search box, so people do not need to change how they search.
One thing to keep in mind throughout: these are the queries we track on the Productrise platform, which are generally product-centric.
Key findings at a glance:
- Standard search returned products on about 88% of queries. AI Mode returned them on just 23%.
- When products did appear, standard search showed about 22.5 products per SERP versus 4.3 in AI Mode, roughly 5.3x fewer.
- Combining both effects, AI Mode surfaced about 95% fewer product listings overall on the same queries.
- Only about 0.8% of products visible in standard search also appeared in AI Mode for the same query the same day.
Why this matters now
Google keeps pushing people toward AI answers. AI Overviews already sit on top of normal results for many searches, AI Mode is rolling out to more countries, and a part of the SEO world now expects AI-first search to become the norm rather than the exception. Because AI Mode opens right from the search box, this shift does not ask users to do anything different.
Google says its AI features are good for the open web and send plenty of traffic to websites. But for online stores, our data showed AI Mode surfacing only a small fraction of the products that standard search does. That means far fewer products get seen, which is why we tracked both surfaces side by side on the same queries.
Finding 1: AI Mode rarely shows products at all
The first gap is whether a shopping search shows any products at all. On the same queries, tracked on the same day, standard search showed products on about 88% of them. AI Mode showed products on only about 23%.
You can see this on a single search for "neck cream". On the left, standard search fills the page with a crowded product carousel you can buy from directly. On the right, AI Mode answers the very same search with a block of text that links out to reviews, without a single shoppable product listing.
For the shopper, that is the difference between a page full of products to buy and a written answer that points to a few reviews instead.
It's still early, but the brands taking the initiative to accurately align and optimize their product data are going to stay three steps ahead. The bigger question I have is whether AI Mode's limited product visibility is a permanent feature or simply a temporary stage as Google expands its monetization model.
Google is already testing sponsored product experiences in AI Mode, but as those placements become more prominent, will it also expand the opportunities for organic product listings? At that point, AI Mode may start to look less like a completely new search experience and more like a visual evolution of traditional search, with a familiar balance of paid and organic results.
That would make product visibility more competitive on both sides. Brands would need to invest more strategically in paid placement while treating product data optimization as a core organic search channel.
J
Finding 2: when products do appear, there are far fewer of them
Even on the few searches where AI Mode did show products, the list was much shorter. Standard search showed about 22.5 products per SERP that returned products, while AI Mode showed about 4.3 per answer. That is roughly 5.3x fewer spots to compete for, and that is before you count how rarely AI Mode shows products at all.
The screenshots below show the same search for "1/2 inch plywood" on both surfaces. Standard search returns a single carousel of ten products, while AI Mode narrows the same search down to just four.
With so few slots, fewer brands make the cut, so the competition for each spot is much tighter in AI Mode.
Starkly fewer products, but a much narrower focus reinforces Google's commitment to helping people find what they need quicker. That's probably my biggest takeaway from this study: Google is repackaging product information in a way that's easier to deliver to the user. Instead of wading through 20+ products that look the same half the time, customers get to the point faster, armed with touch points they may or may not have been thinking about.
It also tells me optimized product information has gone beyond table stakes. A card with a picture and a price isn't enough to entice anyone anymore. If Google is using reviews and other citations to subcategorize products, it's time to ask yourself whether yours is a product worth showing.
J
Finding 3: the combined effect is a 95% collapse
Put the two effects together, far fewer searches with products and fewer products when they do appear, and you get the headline number. Across the same set of queries, AI Mode showed about 95% fewer product listings than standard search.
The combined effect
Total product visibility, AI Mode compared with standard search.
-95%
fewer product listings in AI Mode than standard search
Source: Productrise · productrise.app
A headline number this big is easy to read as a shrinking market, but that is not the only way to look at it. Fewer listings on a single response says nothing about how many chances a brand still gets across a full, back-and-forth shopping journey.
The findings are interesting, but I'd be careful not to interpret them as a collapse in opportunity. AI search fundamentally changes the way people discover products, and those journeys are often conversational rather than a single query. Fewer products per response doesn't necessarily mean fewer opportunities for brands to be recommended over the course of an entire shopping journey.
What's becoming increasingly clear is that Google Shopping optimisation, product data, brand signals and broader organic visibility can no longer be treated as separate disciplines. AI systems need confidence in both the product and the brand before recommending them, so success comes from strengthening the entire ecosystem rather than optimising individual channels in isolation.
It's also worth remembering that AI shopping experiences don't exist in isolation. ChatGPT already uses Google Shopping as one of its sources for product discovery, so improvements to Organic Shopping visibility can extend beyond Google's own interfaces. Optimising for Google Shopping is increasingly becoming an optimisation for the wider AI ecosystem, not just Google.
C
What this means for brands and SEOs
Visibility in product carousels does not automatically mean visibility in AI Mode. Doing well in standard Shopping carousels is no guarantee your products turn up in AI Mode. As the overlap number above shows, the two surfaces pick largely different products, so a strong position on one does not transfer to the other.
Google seems to pick quality over quantity. A standard search page can list as many as 60 products, so even non-optimized products often still find their way onto the first page. That is much less likely to happen in AI Mode. So brands have to focus on getting their product data quality in order, which at the very least means having optimized titles and descriptions for your entire product catalogue.
Matching your product data and content to AI Mode matters more than ever. Google has launched conversational attributes to help brands get their products surfaced in AI Mode. Brands should not sleep on attributes like these.
My best guess for why the product lists are so short is that AI search simply behaves differently. It's conversational, and if Google expects you to ask a follow-up anyway, whether that's cheaper products, a different brand, or faster shipping, it doesn't need to show twenty products upfront.
The flip side is that being one of four or five recommended products is a very different position than being one of twenty. Click-through rates from these AI Mode sessions should be considerably higher for the products that do make the cut. And we've already seen reports that traffic from AI conversations tends to convert better than average. So while the visibility pie is shrinking, each slice is likely worth more. What that nets out to for a brand's bottom line is going to be one of the more interesting questions to watch.
H
Curious how product carousels behave in normal search?
Long before Google started expanding AI Mode into their main products, we were already tracking how organic product carousels show up in standard Google search: how often a search page contains one, and how many appear when it does. If you want the wider context behind the numbers above, these two trackers have been running for a long time and update continuously.
Our methodology
We tracked thousands of product-centric shopping queries, run against both standard Google search and AI Mode on the same day, so every comparison is like-for-like. In total the study drew on more than 2 million product listings across more than 100,000 SERPs and AI Mode responses, over three weeks (21 days) in July 2026, using clean, non-personalized sessions. It covers the US and UK only, where Google's Shopping Graph feeds directly into AI Mode answers, the same engine behind traditional Shopping results. Many other markets do not return rich product results in AI Mode yet, so we left them out.
We compared the same keyword query on both surfaces, which is also how most people reach AI Mode today: it opens from the normal search box, so searchers arrive with ordinary shopping keywords rather than long conversational prompts.
A "product" means one distinct listing: a card in a standard Shopping carousel, or a product inside an AI Mode answer. Averages exclude empty results, and to match products across the two surfaces we normalize titles, since AI Mode shows Google's cleaned names while standard search uses raw feed titles.
About this study
This study was produced by Productrise, a platform for tracking organic product visibility in Google. All figures were computed directly from our tracking database over the study window. You are welcome to reuse, reference, or republish the charts and findings, as long as you attribute Productrise and link back to this page. Questions or data requests are welcome at hello@productrise.app.
