Your search results page is advertising inventory you already own and have never sold. So is your product page, and so is your order confirmation.
Thrive Market launched a retail media network on Carrot Ads in 2024 rather than building an ad stack from scratch. That is the part worth copying: the infrastructure is now rentable, which removes the engineering budget that kept this a big box play. Chewy, by its own account, sees about 1 in 3 ad clicks convert to a purchase, which tells you how well intent converts when the ad sits next to the buy button.
The shape of it is simple. Complementary brands that do not compete with your core line pay for placement inside your search and discovery results, a slot on relevant product pages, and a unit on the post purchase page. A food brand sells placement to a beverage brand. A skincare brand sells to a supplement brand. You serve the ads through an app, report performance back in aggregate, and invoice monthly. The revenue carries no cost of goods, which is the whole point.
The first catch is that media sales is a business, not a feature. Somebody pitches, negotiates, traffics creative, reports, and chases invoices. If nobody owns that job, you will build the placements and sell 2 of them.
The second catch matters more. Every sponsored slot is a slot not serving the most relevant result. Sell too many and your conversion rate quietly pays for your ad revenue. Cap density, protect high intent queries, and watch sitewide conversion as the primary metric, not media income. On reporting, keep it aggregated. Advertisers will ask for customer level data and you should not give it to them.
You have spent a decade paying Meta and Google for access to other people's audiences. Somebody in your category would happily pay for access to yours, and they are already trying to reach those people the expensive way.
