Pinterest Ads and Video: Inside the Platform That Quietly Became a Performance Channel

Pinterest is the strangest social platform to write about. It isn't quite social — most users don't follow people, don't post comments, don't chase virality. It isn't quite a search engine — though more than 8 out of 10 weekly users say they discover new brands through it. It isn't quite e-commerce — though it generates roughly $4.30 in attributed sales for every $1 spent on ads. It is, increasingly, all three at once: a visual search engine sitting on top of a shopping graph, with an ad system optimized for the long tail of buying intent.

For a decade, Pinterest was treated by media planners as an "always-on, low-priority" channel. In 2026, that posture is harder to defend. The platform crossed 600 million monthly active users in Q3 2025, daily video views are up roughly 240% year-over-year, and Pinterest has spent the last 24 months rebuilding its ad stack around AI-driven automation, contextual signals, and CTV expansion. This is a guide to where the platform actually is right now — what the ad ecosystem looks like, why video has taken center stage, how targeting works under the hood, and what's changing in 2026.

What makes Pinterest different from other ad platforms

Three structural facts shape everything else.

Pinterest users arrive with intent. On Meta or TikTok, users come to be entertained; advertising interrupts that. On Pinterest, users come to plan — kitchen renovations, weddings, fall wardrobes, vacation itineraries, gift lists — often weeks or months ahead of purchase. Roughly 85% of weekly users say they've made a purchase based on a Pin. Around 96% of searches are unbranded, meaning the user is looking for a solution, not a brand. That gap is where new advertisers can win share without outbidding incumbents.

Pins compound. A Pin posted today can still drive impressions and clicks three months later. Promoted Pins continue generating organic engagement after paid spend ends — when a user saves a promoted Pin, the "promoted" label disappears, and what was paid media becomes earned reach. Roughly half of ad-driven purchases on Pinterest happen more than two weeks after the initial impression. This breaks most attribution models built for Meta's 7-day click window.

The audience skews differently. Around 70%+ of users are women, Gen Z is the fastest-growing cohort (now~42% of global users), and over 80% of the audience lives outside the United States. ARPU is heavily skewed toward the U.S. ($7.64 in Q3 2025) versus Europe ($1.31) and the rest of the world ($0.21). In other words, international markets remain drastically under-monetized. For brands with non-U.S. footprints, that's an arbitrage window.

The Pinterest ad ecosystem in 2026

Pinterest's ad inventory is now wider and more specialized than most marketers realize. The current format roster — per Pinterest's own ad spec documentation and up-to-date 2026 spec guides:

Standard Pin Ads. Single-image promoted Pins, recommended at 1000×1500 (2:3 aspect ratio). Native, low-friction, and the workhorse of most campaigns.

Video Pin Ads. Available in standard width (matching static Pin dimensions) and max width, which expands to fill the entire mobile feed. Autoplay triggers when 50% of the Pin is in view. Recommended runtime is 6–15 seconds for paid; 45–90 seconds for organic Video Pins. Aspect ratios from 1:1 to 9:16 are supported, with vertical performing best.

Carousel Ads. Up to 5 swipeable images in a single Pin, each with its own title, description, and destination URL. Strong for product ranges and step-by-step storytelling.

Collections Ads. A hero asset (image or video) followed by up to 24 secondary images, expanding into a full-screen mini-shop on tap. Designed for e-commerce catalogs.

Shopping Ads. Product-feed-driven single-image ads with the highest measured ROAS of any Pinterest format — averaging 2.3x against 1.7x for consideration and 1.2x for awareness campaigns. The natural format for retailers with synced catalogs.

Showcase Ads. A title card plus up to 4 swipeable feature cards, each supporting interactive overlays. Good for highlighting multiple products or features within a single creative unit.

Quiz Ads. Interactive title card → up to 3 multiple-choice questions → a personalized results card. Used heavily by beauty, fashion, and home brands to route users to product recommendations.

Idea Ads. The multi-page, full-screen storytelling format evolved from organic Idea Pins. Engagement rates typically 4–8x higher than standard Pins.

Premiere Spotlight. Pinterest's premium reservation product. A max-width video ad that takes the first ad slot on the home feed (page 2) or the first slot in the search carousel for an entire day. Around 50% of mobile screen real estate. Brands report 8–12x higher recall than standard placements; pricing reflects that.

This list grew. Three years ago, Pinterest essentially had Promoted Pins and Promoted Videos. The current spread reflects a platform deliberately trying to give performance advertisers and brand advertisers separate tools, instead of one generic ad unit.

Why video became the headline format

Pinterest's roots are static. So the video push deserves an explanation.

The performance gap is real. Pinterest's own creative benchmarks put video Pin ads at roughly 3.2x the engagement of static Pins, with about 2.4x higher brand lift. Save rates run 0.3–0.5% on video against 0.15–0.25% on static — a meaningful, consistent gap. Beyond the numbers, video answers Pinterest's core search behavior better than a still image: a user searching "small kitchen storage" is more convinced by a 6-second clip showing a drawer organizer in motion than by a glamour shot.

A few practical realities every marketer running Pinterest video should internalize:

• Sound is off. Around 85% of users watch without audio. If your story depends on voiceover, the story fails. Use captions, on-screen text, and visual hooks.
• You have ~1.7 seconds. That's the median scroll-stop window on Pinterest. The hook, the brand, and the value proposition need to land inside the first 3 seconds.
• Vertical wins. 9:16 and 2:3 outperform 1:1, which outperforms 16:9. Mobile-optimized vertical creative has been shown to drive completion rates up to 80% higher than horizontal across visual discovery platforms.
• 6–15 seconds is the paid sweet spot. Longer runs better organically, but paid placements perform best in the bumper-to-short-form window.
• Branding stays present. Show the logo or product within the first 3 seconds and hold it at the end. Pinterest's autoplay-then-scroll pattern means many viewers never reach the back half.

How targeting actually works on Pinterest

Pinterest's targeting stack has three layers that most advertisers think about, and a fourth that's becoming the differentiator.

Keyword targeting maps ads to user search terms. Because Pinterest is a visual search engine more than a feed, this layer is unusually powerful — closer in mechanics to Google Ads than to Meta. Match types matter; broad match on Pinterest tends to leak into adjacent intent that may or may not convert.

Interest targeting uses Pinterest's taxonomy of categories (home, beauty, food, etc.) and is most useful for top-of-funnel reach. It's blunt by design; performance campaigns rarely lean on it.

Audience targeting includes customer list match, retargeting via the Pinterest Tag, engagement audiences (people who've interacted with your Pins), and "actalikes" — Pinterest's lookalike equivalent, modeled on behavioral affinities rather than declared interests.

Contextual targeting is the layer that's been quietly transforming. Cookieless by definition, contextual targeting places ads based on what's in the surrounding content rather than who the user is. On a platform like Pinterest, where most engagement is content-driven rather than identity-driven, contextual signals can be more predictive than user-level data — and they remain stable even as third-party identifiers continue to degrade.

For static Pins, contextual targeting is relatively well understood: it operates on image classification, board context, and keyword adjacency. For video Pins, it's harder. Conventional contextual systems tag a video by metadata, title, and surface-level keywords — which means a yoga mat ad placed against "morning routine" videos may end up running against everything from skincare hauls to commute vlogs. The signal is too coarse for performance advertisers in regulated or specialized verticals.

This is where AI-powered video contextual platforms have started to fill the gap. PXLSTRM, the AI contextual targeting spinoff of Swiss/US AdTech firm Adello, is one of the better-known examples in the European market: its AI analyzes video content at the object, scene, and dialogue level — not just metadata — and clusters millions of analyzed videos by genuine semantic similarity. 

The platform launched on YouTube in 2023, expanded to TikTok, then into Asia-Pacific, and as of March 2026, supports Pinterest campaigns across the DACH region (Germany, Austria, Switzerland), with reach to roughly 24 million high-intent users in those markets. Reported relevance uplift across markets averages above 100%, with comparative tests against YouTube TrueView showing meaningful media savings while reaching more on-topic audiences. For advertisers in regulated categories (alcohol, pharma, gambling, finance) — a category set that matters disproportionately in DACH — contextual video AI is rapidly becoming the path of least resistance to brand-safe scale.

The broader point: in 2026, "targeting on Pinterest" no longer means just keywords plus interests. It means a stack — first-party platform signals, behavioral audiences, plus a contextual video layer.

Pinterest's own AI: Performance+

Pinterest is also automating its way up the stack. Performance+ is the company's AI-driven campaign suite, comparable in spirit to Meta's Advantage+ or Google's Performance Max. It bundles automated bidding, creative optimization, and audience targeting under a single objective.

The numbers Pinterest reports on Performance+ adoption are strong: retail advertisers' spending on Performance+ saw an average 24% higher conversion lift, and mid-market and small advertisers saw meaningfully higher monthly spend growth post-adoption. CPA across Pinterest more broadly dropped roughly 28% year-over-year in Q4 2025, with Performance+ delivering ~18% lower CPA than manually optimized campaigns.

The trade-off is the same as on every automated platform: less manual control, less granular reporting, and a creative-quality dependency. Performance+ rewards advertisers with clean catalogs, multiple creative variants, and well-instrumented conversion tracking. Without those, the AI has nothing to optimize against, and the budget gets wasted on cheap but irrelevant impressions.

Measurement, attribution, and the long-tail problem

The Pinterest Tag — a JavaScript pixel — handles conversion tracking, retargeting audiences, and catalog measurement. The server-side conversion API is also available and increasingly recommended for accurate measurement under iOS privacy constraints.

The harder problem is attribution windows. Pinterest's purchase journey is long: as noted, roughly 50% of ad-driven purchases happen more than two weeks after first exposure. A 7-day click model — the default in most last-click systems — undercounts Pinterest dramatically. Marketers running unified MMM (marketing mix modeling) or MTA (multi-touch attribution) tend to discover that Pinterest's true contribution is significantly higher than last-click models suggest, particularly in home, fashion, and beauty.

Practically, this means Pinterest deserves longer attribution windows (28-day click, 1–7 day view-through at minimum) and incrementality testing rather than naive last-click ROAS comparisons.

What's coming in 2026 and beyond

Three structural shifts are worth tracking.

CTV expansion via tvScientific. Pinterest closed its acquisition of CTV ad-platform tvScientific in February 2026 and launched tvScientific by Pinterest shortly after, extending Pinterest's first-party audiences into connected TV inventory. Early test data from LG showed a 73% increase in unique households reached and a 24% lift in net new customers when Pinterest audiences were layered onto CTV. This effectively turns Pinterest into a cross-screen platform for the first time.

AI shopping agents. During Cyber Week 2025, around 20% of global e-commerce orders were influenced by AI shopping agents — an inflection point. Pinterest's visual search and product graph make it unusually well-positioned to be a source surface for those agents, which has implications for how brands structure product feeds and Pin metadata.

Video is the dominant ad format. Premiere Spotlight expansion, max-width video defaults, video shopping ads converting at higher rates than standard video, and 240% YoY video view growth all point in the same direction. Brands that haven't yet built a vertical-video pipeline for Pinterest will be at a structural disadvantage by the end of 2026.

The takeaway for advertisers

Pinterest in 2026 is not the platform marketers wrote off in 2019. It's a search-driven, intent-rich, increasingly AI-automated ad system with a long-tail attribution profile and an unusually strong fit for video — particularly contextual, brand-safe video. The advertisers winning on it are the ones treating it like a hybrid of Google Search and YouTube, not like a third Meta channel.

For non-U.S. markets, the structural under-monetization of European and rest-of-world inventory creates a real efficiency window — one that closes as more advertisers move in. The fundamentals (lower CPMs, less competition, an intent-rich audience, growing Gen Z share) won't last forever. They're priced into the next 12–24 months, not the next 5 years.

If your brand is investing in video and not yet running it on Pinterest, the question isn't whether to test — it's how fast you can get a vertical-first creative pipeline live.