Can ChatGPT ad carousels crowd out smaller brands?

What is ChatGPT testing with two ad carousels from one brand?

ChatGPT is testing a layout that shows two ad carousels from the same brand under a single answer. One advertiser can now fill a lot more of the sponsored space than before. The test builds on an earlier experiment that grouped several product feed ads from one retailer into a single unit at the bottom of the response. In that unit, the store's name and favicon sat above product cards showing names and prices. Every card links back to the same website, so any click on any card sends the shopper to that one brand. Advertisers love that, and for the store that didn't buy the placement, it's a quiet problem.

Some context helps here, because ChatGPT ads have moved quickly. The first format was a simple text ad in February 2026. Single product feed ads followed in May, and in August the multi-product carousel appeared, typically showing about five products from one retailer. Big names such as Walmart, L.L.Bean and VistaPrint were spotted in it early. Separately, a doubled text ad format featuring two ads for the same domain has also been seen in testing. The direction is clear. Each new test gives one advertiser more room, and the platform, not the advertiser, decides which format appears for a given conversation.

If you want the details of the original single-brand unit, including where it sits and what it shows, we've covered how single-brand shopping ads affect organic product listings separately. What matters for this article is the trend line. Two carousels from one company means up to roughly ten paid product slots pointing to a single store, sitting right under an answer a shopper has just read. For a smaller retailer, clinic or service business, that space is effectively priced out of reach. The obvious question is whether anything is left for everyone else, and there's more left than you'd expect.

How much of a ChatGPT shopping answer stays organic?

The answer itself, the part that names products, explains trade-offs and makes a recommendation, stays organic. Sponsored units are shown separately and labeled, and they sit at the bottom of the response rather than inside it. That split matters more than it might seem. A shopper who asks for the best running shoes for flat feet reads the recommendation first. The carousel arrives after they've formed a view, and a brand mentioned in the answer has already been endorsed in the assistant's own voice. An ad simply can't buy that kind of credibility, no matter how many cards it holds.

That doesn't make the ads irrelevant. Two carousels from one brand take up real screen space, especially on mobile, where a shopper may scroll past the organic text and land on product images with prices attached. Images pull the eye. So the honest picture is that the paid area is getting bigger and more concentrated, while the organic answer remains the part that shapes the decision. Smaller businesses should stop asking how to match the carousel and start asking how to get named in the answer above it. Naming the right brands for the right reasons is the assistant's whole job, so that's where the real competition is.

Organic product results are also more than plain text. ChatGPT can show its own product cards and comparison layouts in the unpaid part of the answer, built from merchant product data and what it finds on the open web. Those cards are chosen on relevance and quality signals, not bids. We've looked at where clicks from ChatGPT's Shop now buttons actually land and how to track that traffic. For most stores, earning a place among those organic results is a far better use of effort than trying to win an auction against a retailer with a holiday budget a hundred times bigger than yours.

Why is outspending big brands in ChatGPT ads a losing game?

Outspending a big brand in ChatGPT ads is a losing game because the format increasingly rewards concentration. If one advertiser can hold two carousels, the auction favors whoever can pay to fill both. National retailers come with huge catalogs, years of conversion data and seasonal budgets built for exactly this kind of placement. A local bike shop or a niche skincare store can't match that, and trying usually means paying premium prices for the leftover slots on queries the big players didn't want. That's not a strategy. It's a slow leak in your marketing budget.

There's a deeper problem too. Paid placements in an AI answer behave just like paid search ads. The day you stop paying, you disappear. Nothing builds up, no reputation carries over to the next conversation, and there's no asset to show for months of spend. We'd go further than most marketers on this point. For a small business, a ChatGPT ad budget is mostly rented attention in a channel where the landlord keeps adding units for the biggest tenant. Some testing makes sense if you sell high-margin products with a clear conversion path. Making it your main plan for AI visibility doesn't.

The alternative compounds. A product page or guide that ChatGPT cites this month will often still be cited next quarter, as long as it stays accurate and useful. Every review you earn, every comparison page you publish and every feed error you fix adds to a base the assistant draws on again and again. If you're weighing the economics of the two routes, our 12-month cost-per-lead comparison of organic content and paid search walks through when organic overtakes ad spend. The same logic applies inside AI answers, only more sharply, because the paid space there is smaller and fills up faster.

What earns a product or brand a citation in ChatGPT's organic answer?

ChatGPT cites products and brands it can verify and match to the question. In practice that comes down to four things: a clean, complete product feed, pages that compare options honestly, a visible body of real reviews, and content that answers the question in its first lines. Start with the feed. Merchants can submit structured product data, and the basics matter: accurate titles, descriptions that say what the product is for, correct brand and seller names, working image links, current price and real availability. Feeds go stale fast, so refresh yours regularly. An out-of-stock item or a wrong price is an easy reason for an assistant to pick someone else.

Comparison pages are the most underrated asset here. Shoppers ask ChatGPT things like which espresso machine is better for a small kitchen, or whether a standing desk is worth it for someone under five foot five. A page that compares your product with real alternatives, competitors included, and says plainly who each option suits gives the assistant something it can quote. Service businesses can do the same thing. A law firm can compare flat-fee and hourly billing, and a clinic can explain when one treatment beats another. Being fair is what makes it work. A comparison that always crowns you the winner reads like an ad, and assistants treat it like one.

Reviews and answer-first content complete the picture. Assistants pick up sentiment from reviews on your site and across the web, so a steady stream of specific, recent customer feedback beats a handful of five-star blurbs. On your own pages, put the direct answer in the opening sentence, then back it up with specifics such as sizes, materials, timelines and what's included. Use clear headings that mirror real questions, and make sure AI crawlers can actually reach your pages. Want a quick read on how citable your content looks right now? Our free GEO score checker flags the gaps worth fixing first.

How can smaller businesses compete with big brands in ChatGPT?

Smaller businesses compete in ChatGPT by being the best answer to narrower questions, not by chasing the broad ones big brands dominate. A national retailer will probably win the paid space for a query like "best running shoes". It's much less likely to have the best page on running shoes for wide feet after a bunion operation, or the clearest guide to choosing a family dentist who treats anxious children in a particular town. Specific, high-intent questions are where AI assistants go looking for specialist sources, and where a small team with real expertise beats a bigger company's generic category page.

Focus is what turns that into a plan. Pick the twenty or thirty questions your best customers actually ask before they buy, the ones you hear on the phone, in your inbox and at the counter. Write one strong page per question, answer it in the first lines, compare honestly and keep the facts current. Fix your feed so every product the assistant might recommend has complete, accurate data. Then build out reviews where your buyers already leave them. None of this needs an agency-sized team, but it does take consistency, and consistency is what most small businesses run out of time for.

That's the gap we built Artiql to close. It works as an organic-marketing autopilot, producing answer-first, citation-ready content so your site ranks on Google and Bing and gets named in ChatGPT, Gemini, Claude and Perplexity, without hiring a content team. Pricing is tailored to each business after a short discovery call, so there's no one-size plan to compare. If big brands are about to fill the carousels in your category, the useful move is to make sure you're the source the answer itself relies on. To see what that would look like for your site, book a short demo and we'll walk through your highest-value questions together.

Frequently asked questions