Pinterest is pushing deeper into performance advertising with a new format designed around visual search.
At Pinterest Presents on September 17, 2026, the company announced Visual Search Ads, a lower-funnel ad format that can place advertisers in Pinterest Search Results and Pin closeups.
The format is designed for an unusual part of the buying journey:
the moment when a user can recognize what they want visually but may not yet have the exact words to describe it.
Pinterest says users conduct more than 80 billion searches per month on the platform, with the vast majority being visual and more than half commercial.
That claim comes from Pinterest’s own data and should be treated as a platform-reported statistic.
The strategic point is still important: visual discovery is becoming a monetizable search surface.
Visual Search Ads target intent before the query is fully verbalized
Traditional search advertising begins with words.
A user types:
“brown leather tote.”
Visual search can begin with an image.
A user opens a Pin, identifies a bag visually and starts comparing related products.
Pinterest’s new ad format lets brands bid for a prominent position in that discovery flow.
According to Pinterest, advertisers can appear:
- in search results;
- in Pin closeups.
The ad can send users directly to the advertiser’s site.
This makes the visual object itself part of the query.
Pinterest is trying to own the pre-decision moment
Pinterest CEO Bill Ready described the opportunity as reaching people while they search, compare and move closer to a decision.
The platform has spent years positioning itself differently from traditional social feeds.
The argument is that Pinterest users are not only consuming content.
They are planning:
- purchases;
- rooms;
- outfits;
- weddings;
- meals;
- trips;
- projects.
That gives the platform a potentially valuable commercial position between inspiration and transaction.
Visual Search Ads attempt to convert that position into direct performance inventory.
The format works with Performance+
Pinterest says Visual Search Ads are compatible with Pinterest Performance+, its AI-powered performance suite.
The company is also adding new Performance+ controls.
Priority Products
Advertisers can tell the system to emphasize selected products during:
- seasonal launches;
- promotions;
- category pushes.
This is a useful response to one of the biggest complaints around automated advertising:
automation can optimize efficiently while ignoring merchandising priorities.
Priority Products adds a business constraint.
The model still automates delivery.
The operator can steer what deserves emphasis.
Pinterest is also adding self-serve A/B testing
Pinterest announced self-serve A/B testing across:
- creative;
- targeting;
- Performance+.
That matters because automation makes causal validation more important.
If an advertiser changes:
- bidding;
- creative;
- targeting;
- product priority;
at the same time, observed performance cannot identify which change mattered.
Built-in A/B testing makes controlled comparison easier.
It also signals that Pinterest wants to compete not only for media spend but for performance-measurement credibility.
App Promotion is moving into global beta
Pinterest also announced App Promotion, an objective designed around app installs.
The feature is in global beta for qualified advertisers.
Pinterest cited one advertiser seeing up to an 85% improvement in cost per install compared with previous workarounds.
That is a platform-supplied case and not a universal benchmark.
The broader implication is more useful:
Pinterest is expanding the number of campaign objectives it can optimize natively.
That moves it further from awareness inventory toward full-funnel media.
Pinterest Intelligence is the infrastructure underneath
Pinterest describes Pinterest Intelligence as a combination of:
- first-party signals;
- visual understanding;
- recommendation systems;
- purpose-built AI models.
The system is intended to interpret not only what users search for but how their preferences evolve as they browse.
For advertisers, the promise is that visual and behavioral signals can be converted into commercial intent.
That makes visual search different from a standard keyword auction.
The signal includes:
- object;
- style;
- context;
- taste;
- behavior.
What changes for performance teams
Creative becomes query infrastructure
On a traditional search engine, the keyword is explicit.
In visual search, product imagery can influence whether the product is recognized as relevant.
That increases the importance of:
- clean product images;
- multiple angles;
- accurate product feeds;
- strong category metadata.
Product feeds matter more
Performance advertising becomes stronger when visual inventory can connect directly to structured product data.
Advertisers should verify:
- product title;
- category;
- availability;
- price;
- destination URL.
Merchandising enters media buying
Priority Products means media teams need closer coordination with:
- ecommerce;
- inventory;
- promotion calendars;
- merchandising.
Incremental measurement matters
Pinterest users are often already in discovery mode.
Advertisers should ask whether ads create additional conversions or capture conversions that would have occurred organically.
A/B testing and lift measurement become important.
Who should test Visual Search Ads first
The obvious categories are visually driven.
Examples:
- fashion;
- beauty;
- home;
- furniture;
- design;
- food;
- travel;
- automotive accessories.
The format is likely to be most useful where the visual object contains information that text does not communicate efficiently.
A user may not know the design name of a chair.
They can still know exactly which chair they want when they see it.
What to measure
Do not stop at clicks.
Measure:
Discovery
- visual-search impressions;
- qualified clicks;
- product engagement.
Commerce
- add-to-cart;
- checkout;
- purchase;
- new customer rate.
Efficiency
- CPA;
- ROAS;
- marginal ROAS.
Incrementality
- lift vs. control;
- new-to-brand behavior;
- assisted conversion.
The strategic implication
Search is becoming multimodal.
The keyword remains important.
But commercial intent can increasingly appear through:
- image;
- video;
- conversation;
- voice.
Pinterest is positioning itself to monetize one of those modes directly.
For marketers, Visual Search Ads are not simply another placement.
They are evidence that product discovery is becoming less dependent on the user translating desire into a typed query.
The platforms that can understand the object, context and commercial intent at the same time gain another opportunity to mediate the purchase journey.
A test plan for Visual Search Ads
Advertisers entering the beta should avoid evaluating the format as a simple new placement.
Build a test around a specific hypothesis.
Example:
Visual Search Ads will acquire new customers at a comparable contribution margin to standard Pinterest performance campaigns because users are closer to a product decision when they expand or visually search a Pin.
Track:
- new-customer rate;
- product-view rate;
- add-to-cart rate;
- conversion;
- gross-margin ROAS.
Use the self-serve A/B testing capability where available.
Creative setup
Prepare:
- high-quality product images;
- lifestyle imagery;
- multiple visual contexts.
Feed setup
Validate:
- product identifiers;
- price;
- availability;
- destination pages.
Measurement
Separate:
- attributed conversions;
- new-to-brand;
- incremental outcomes where possible.
The competitive question
Pinterest is not only competing with social platforms.
Visual Search Ads move it closer to:
- search engines;
- retail media;
- commerce discovery platforms.
The competitive advantage Pinterest is trying to monetize is intent before language.
A shopper may not know the name of the interior-design style they want.
A visual system can still infer similarity and commercial relevance.
If the product performs, advertisers gain another lower-funnel surface.
If it does not, the lesson may be that visual inspiration does not automatically translate into incremental purchase intent.
That is exactly what the new testing tools should help determine.
What would make the format strategically important
Visual Search Ads become more than a tactical test if advertisers can repeatedly prove three things:
- visual-discovery traffic converts at a strong rate;
- the format adds new demand rather than only intercepting existing Pinterest intent;
- product-feed and merchandising controls let businesses steer automation without destroying efficiency.
If those conditions hold, Pinterest gains a differentiated performance surface rather than another social ad placement.
That matters because media buyers are increasingly looking for inventory that sits closer to commercial intent while still offering discovery.
Visual search occupies exactly that boundary.
The next question is not whether the format can generate clicks.
It is whether visual intent can become a scalable, incremental acquisition signal.
