Pinterest Put Visual Search Ads Where Shoppers Make Up Their Minds

Pinterest introduced Visual Search Ads on September 17, putting a sponsored result inside the image-led searches people use when they’re comparing products. The ad can appear in Pinterest Search Results and Pin closeups, where the company says a shopper is still deciding what they want.

That placement matters because visual search starts before a customer has the language for a product. Someone can begin with a handbag, a countertop, a dress, or a room they like, then keep narrowing the idea through images. Keyword search still has a role in that journey. It just isn’t the first signal anymore.

Pinterest says people conduct more than 80 billion searches on the service each month and that more than half are commercial. That’s Pinterest’s own figure, not an independent measure. But it explains why the company sees a performance opportunity in a format that sits beside an image rather than below a typed query.

Bill Ready, Pinterest’s CEO, framed the move as a response to richer search behavior:

“Traditional search advertising has long relied on keywords, but in an AI-driven world, people are expressing themselves in much richer ways.”

Visual Search Ads are available in beta to eligible advertisers in Pinterest’s ad markets in the coming weeks. Pinterest says advertisers can bid for a top response and send people to their own site. The company hasn’t published conversion rates, cost-per-acquisition results, or a study comparing the new unit with ordinary Pinterest search ads.

A new placement is not a new measurement plan

The launch comes with two related changes to Pinterest Performance+, the company’s AI-assisted campaign suite. Priority Products lets advertisers tell the system which products matter most during a seasonal launch, promotion, or category push. Pinterest also announced self-serve A/B testing across creative, targeting, and Performance+ campaigns.

The new controls are useful because automation creates a familiar tradeoff. Performance+ can simplify setup, automatic bidding, and creative production. Its documentation says it can generate product-image backgrounds, resize images, and turn catalog product Pins into Shopping and Collections ads. In return, the marketer has less direct control over some delivery choices, including targeting and how spend moves between automatically created ad groups.

That makes the test design more important than the launch itself. A marketer should decide before spending whether a visual-search placement is supposed to improve discovery, qualified site traffic, add-to-cart rate, or purchase efficiency. Those are different jobs. Treating a top-of-funnel image click as proof of lower-funnel performance would make the result look better than it is.

Pinterest’s own guidance says Performance+ Lead and Sales campaigns work best with 50 or more weekly conversion events. It also recommends a daily budget at least five times the target cost per result for teams optimizing toward a CPA goal. Those are platform recommendations, not a guarantee of performance. They are still a useful warning that a lightly funded visual-search test may not produce a clean answer.

The product page becomes part of the creative

Visual search also changes what counts as ad creative. The image that pulls someone in may not be an advertiser’s asset. A shopper could start with an inspiration Pin, then see an ad for a similar product. The landing page has to close the gap between the visual idea and the product being sold.

That puts more pressure on product photography, feed accuracy, availability, color and variant information, and the first screen of the landing page. A campaign can find a person who likes the shape, texture, or setting in an image. It can’t fix a product page that makes the shopper wonder whether the item is the same one they thought they found.

For brands with a strong catalog and visual category fit, Visual Search Ads are worth a controlled beta test. Start with a narrow product group and a clear conversion event. Keep a comparable campaign or holdout where possible. Review the search context, the asset, the landing page, and the conversion path together. Pinterest positions itself as a place where people discover, plan, and shop, but that doesn’t answer whether a given ad placement adds incremental customers.

This also belongs beside the broader AI-shopping traffic question, not apart from it. A visual ad may influence what people discover before they use a search assistant or branded search to finish the purchase. It also looks different from the workflow problem in ChatGPT Ads inside HubSpot and Shopify, where the key question is whether the campaign signal makes it back to the customer record.

Pinterest has made visual intent more buyable. It hasn’t made it self-explanatory. The teams that learn from the format will be the ones that separate a new placement from a real incremental result.