The brand ChatGPT recommends is almost never the one advertising, Floodlight data shows

OpenAI opened its self-serve ad platform to all advertisers in May 2026, dropping the entry point from a $200,000 managed-service minimum to open access. The early coverage was predictable: Pilot brands like Expedia, Ford, Target, and Adobe got named in headlines, the platform reportedly hit $100 million in annualized revenue within its first six weeks, and the general consensus was that inventory was filling up fast. Floodlight, a programmatic ad solutions provider, ran 550 live queries across 11 consumer product categories to see if that was actually true. Here's what the results showed.

TL;DR: 84% of the 550 queries Floodlight tested returned zero ads, but in the verticals where ads did appear, the brands claiming those slots mostly had nothing to do with the product being asked about.

Testing Methodology

All 550 prompts ran on a free-tier ChatGPT account between May 18 and June 4, 2026. Currently, OpenAI only serves ads to non-paying users, so a paid subscription would have returned zero results regardless of what was being asked. The 11 verticals tested included baby gear, baby monitors, baby carriers, stroller wagons, car seats, cribs and crib mattresses, mattresses, beach tents, dog food, arch support sneakers, as well as women's care and clean beauty. Floodlight included 50 prompts per vertical.

Each vertical's prompts were written across four intent types: buying intent, brand-focused, use-case, and informational. More importantly, they were written the way a real person actually types into ChatGPT. Nobody opens ChatGPT and types "arch support sneakers." That's how they'd use a search engine like Google. Instead, they type something like: "I've been dealing with plantar fasciitis for six months and my podiatrist told me to get better arch support, what sneakers should I be looking at?" That's the kind of prompt that got used.

Each of these ads was confirmed via DOM inspection and taking a screenshot. If it didn't show up in the website's code, it didn't count. Keep in mind that this is three weeks of data on a platform that opened to everyone about two months ago, so treat this data as a first look.

84% of Queries Returned Nothing

Of the 550 queries, 464 returned no ad at all. That's 84.4% without a sponsored result, brand name, or anything below the answer. Instead, the platform shows ChatGPT's organic response and a blank space where an ad could have been.

There are five categories that had zero ads across every single query: beach tents, dog food, arch support sneakers, women's care and clean beauty, and car seats. To put this into perspective, a user asking ChatGPT which dog food is best for a senior lab with joint issues got a thorough answer and zero ads.

This same result happened for every car seat query, every beach tent query, and every clean beauty query in the set. That's five categories where no brand has claimed any ad space (yet) and where any brand that sells those products could be the first to show up in these placements.

However, three categories had a bit of activity. Cribs and crib mattresses had 80% of queries claimed, mattresses came in at 38%, and baby monitors only had the brand Nanit present. Everything else was somewhere between thin and empty.

The Wrong Brands Are Claiming the Right Queries

Off-Category Brands Are Showing Up on High-Intent Slots

When someone searches for a specific product they're ready to buy, there should be an ad from a brand that sells that product. But as of writing, that's not the case in many categories.

Instead, the ad space is getting taken up by completely unrelated brands (hotels, luggage companies, grocery delivery apps, department stores). This isn't because they're running a genius cross-industry strategy, but simply because they showed up and nobody else did.

Floodlight's findings showed that upon searching for baby carriers, the ads that appeared were from L.L. Bean and a sporting goods store. In ChatGPT's organic answers, it consistently recommended Ergobaby. Since they hadn't purchased ad space, they're giving competitors the opportunity to start showing up as suggested items.

The same results happened across other products. Stroller wagon searches returned ads for Away luggage and a baby registry site. Women's care products showed Instacart ads. And searches for baby gear showed ads for Choice Hotels, Nissan, Car and Driver.

The takeaway is pretty simple: If you're not advertising on searches for your own product, someone you are in competition with or completely unrelated to your business could collect that customer instead.

The Gap Between Organic Recommendations and Paid Ads

ChatGPT's organic answers consistently recommend the same brands, and these are usually the ones that have great brand recognition, are producing quality insights and editorial content and have glowing reviews:

• In baby carrier queries, Ergobaby came up 26 times.
• In arch support queries, Brooks appeared 20 times.
• In clean beauty queries, ILIA showed up repeatedly.

These brands are earning recommendations and none of them bought the ad unit sitting directly below their own name. By ignoring paid ads, these brands are giving their competitors the opportunity to get in front of their prospective customers.

So the actual experience for a shopper looks like this: They ask ChatGPT about the best baby carriers, ChatGPT says Ergobaby in its organic answer, but then the first ad they see is for L.L. Bean. Ergobaby did the hard work of optimizing their online presence and building a reputation good enough to get recommended by AI, while L.L. Bean just happened to have an ad running.

Ergobaby earns the recommendation. L.L. Bean gets the ad slot beneath it. The pattern stays true across every category where off-category brands appeared. The brand ChatGPT recommends isn't the one paying to show up below it. By all available evidence, they just don't know they can.

The Advantage of Moving Early

The opportunity for brands to capitalize is now. Crib and baby mattress brands have started buying these ad slots, and now 80% of that space is taken. This signals that the inventory doesn't stay empty forever. It fills up the moment the right brands catch on and decide they want to show up.

Brands selling products like strollers, dog food, car seats, and more haven't quite figured that out yet. So right now, the opportunity to capture intent for those queries is there for the taking. Adopting a new channel early often means lower ad costs, less competition, and cleaner data on what's actually converting before everyone else arrives to the party.

For any brand that's already getting recommended by ChatGPT, the math is pretty simple. Recommended brands did the work to build something valuable enough that AI mentions you by name. But someone else is buying the ad that sits right below your name.

The headlines have called ChatGPT advertising a gold rush, but three weeks of data and testing across 550 queries tell a slightly different story. Floodlight found that most categories are empty, and the brands that did show up wandered in because nobody else was there.

This story was produced by Floodlight and reviewed and distributed by Stacker.

Copyright 2026 Stacker Media, LLC

This story was originally published June 22, 2026 at 7:30 AM.