Google’s AI Mode Shows 4 Products, Not 28. Will Yours Make the Cut This Holiday Season?

Black Friday lands on November 27 this year, Cyber Monday on the 30th, and the way a lot of shoppers start that hunt has quietly changed. Instead of typing “wool throw blanket” and scrolling a carousel of thirty options, more people are asking Google a whole sentence — “a warm throw for a reading chair that won’t pill, under $80” — and getting an answer back in AI Mode. That answer doesn’t look like a shelf. It looks like a recommendation. And a recommendation only has room for a few names.

The number that should change your Q4 plan

In September, Productrise published a study that ran the same shopping queries through Google’s classic results and through AI Mode on the same day, from August 9 to 31, 2026 — more than 2 million product listings across over 100,000 results pages in the US and the UK. The headline findings:

• The classic Shopping carousel showed an average of 27.8 products per query. AI Mode showed 3.9.
• Only 1.28% of the products in the carousel also appeared in AI Mode for the same query on the same day.
• When the same product did appear in both, the first-listed seller was different 49.6% of the time.
• When the price differed between the two, AI Mode showed the higher price 68.4% of the time.

It is one third-party study, not Google’s own data, and AI Mode changes fast. But the direction is hard to argue with: ranking well in the Shopping carousel no longer means a shopper asking AI Mode will ever see you. They are two different shelves, and the second one is about seven times shorter.

Why the shelf got so short

A carousel is built to show options and let the shopper choose. A conversational answer is built to do some of the choosing for them — narrow the field, explain why, and suggest a few products that fit what was actually asked. Both draw on the product information Google already holds about your store: your Merchant Center feed, your product pages, your reviews. They just pick from it very differently. A product that wins the carousel on price and bidding can lose the AI answer because nothing on its page says it is a good fit for “won’t pill” or “for a reading chair.”

What actually helps a product get picked

Nobody outside Google has the selection logic, and anyone who promises a guaranteed spot in AI Mode is guessing. What you can control is how easy it is for a system to understand exactly what you sell and who it is for. In our experience, the stores that show up in AI answers have a few things in common:

• Product descriptions that answer real questions, not just a spec table. Who is it for, what is it good at, what does it pair with, how does it hold up after a few washes? That is the language people now type into Google.
• A complete, clean feed. Brand, GTIN, size, colour, material, shipping cost and return window all filled in. Missing attributes are missing reasons to recommend you.
• Price and stock that match everywhere. If your feed says $64 and your product page says $72, you are not just losing a sale; you are giving Google a reason to trust your data less. Fix mismatches before peak traffic, not during it.
• Reviews on the product page itself. A shopper asking “which one is actually good” is asking for social proof, and so is the system answering them.
• Structured data that matches the page. Product, offer, shipping and return-policy markup tell a machine in plain terms what a human sees on screen.
• A page that loads fast on a phone. Every recommendation ends with a click to your site. If that page stalls, the recommendation was wasted.

Notice that none of this is a trick. It is the same work that makes a product page convert a human, which is exactly why it tends to hold up when the algorithm shifts again. If you have read our guide to generative engine optimization, this is that idea applied to a product catalogue.

About that price gap

The study’s pricing finding will tempt some stores to slash prices to “win” AI Mode. We would not. A system that more often showed the pricier listing is not hunting for the cheapest option; it is hunting for the one that clearly fits the question. Make your value obvious instead — what is included, the warranty, free returns, delivery dates before the holidays — and say it in words on the page, not only in a badge image a machine cannot read.

Don’t forget the chat window

AI Mode is not the only conversation happening on Google. In January, Google launched Business Agent, a chat experience that lets shoppers talk to a brand directly from Search — a virtual sales associate answering product questions in the brand’s own voice. It started with large US retailers and eligible stores can switch it on through Merchant Center. Whether or not you turn it on this year, it runs on the same raw material as everything above: the product information you have given Google. Thin data makes a thin salesperson.

A checklist for the weeks before Thanksgiving

• Now: open Merchant Center, clear every disapproval and warning, and fill in missing attributes on your best sellers first.
• Mid-October: rewrite the descriptions of your top 20 products around the questions customers actually ask you — your inbox and DMs are the best keyword research you own.
• Late October: check that price, stock, shipping and returns match between your feed, your product pages and your structured data.
• Early November: test the whole path on a phone — search, product page, cart, checkout — and fix whatever slows it down.
• Thanksgiving week: stop tinkering. Freeze changes to feeds and templates so nothing breaks while the traffic is real.

The bottom line

This holiday season, a lot of shoppers will meet your product inside an answer rather than a list. You cannot buy your way onto a four-product shelf the way you could bid your way up a carousel, but you can make your products the easiest ones on the internet to understand and recommend. That work pays off in AI Mode, in the classic results, and on your own product pages all at once — and if you would like a second pair of eyes on your store before November, that is exactly what our e-commerce team does.