Pinterest yesterday published its first-quarter 2026 earnings, reporting $1.008 billion in revenue - up 18% year over year - and 631 million monthly active users. Those headline numbers matter as context, but the more consequential content for advertising practitioners is buried in the details: a new A/B testing tool in Ads Manager, a retrained ROAS model showing up to 11% gains in experimentation, direct integration of AI bidding with advertisers' own measurement sources of truth, a Canvas image generation model now capable of real-time high-fidelity edits, and CTV audience extension via the newly closed tvScientific acquisition showing a 159% increase in incremental sales in early testing.

This article unpacks those platform changes in the order that an advertiser would encounter them - from campaign setup through to measurement and attribution.

Performance+ campaigns: 30% of lower-funnel revenue and accelerating

Pinterest Performance+ is now the platform's primary automated campaign product, combining automated bidding, budgeting, targeting, and creative into a single workflow. According to CEO Bill Ready on the earnings call, approximately 30% of lower-funnel revenue is now flowing through Performance+ campaigns - a figure reached roughly one year after the full global launch of the product in October 2024.

What stands out in the Q1 data is not just adoption but the spending differential. Advertisers who adopted Performance+ grew their lower-funnel spend at nearly twice the rate of non-adopters in Q1 2026, according to Ready. That gap suggests Performance+ is not simply a migration path for existing budgets - it appears to be pulling incremental lower-funnel investment. Advertisers using Performance+ campaigns continue to see higher ROAS and improvements in both CPA and CPC compared with standard campaigns, according to the press release.

The setup mechanics are designed to lower barriers. Performance+ requires approximately half as many inputs to configure as a standard campaign - a detail that matters less for large teams with dedicated trafficking staff and more for mid-market or SMB advertisers operating without deep operational resources.

Mejuri, a fine jewelry brand, provided the specific case study cited on the call. The brand ran a four-week A/B test in Q1 comparing a dedicated Performance+ campaign to its standard approach. The Performance+ campaign delivered a 46% increase in ROAS and a 62% increase in conversions, according to Ready's prepared remarks. The scale and category - fine jewelry, a considered purchase - is a deliberate signal: Pinterest is making the case that Performance+ works beyond the core retail categories that have historically driven the platform's revenue.

Ready framed Performance+ explicitly as a multiyear product cycle, with "a lot more build still in front of us." The 30% lower-funnel revenue share is the starting point, not the target.

Native A/B testing arrives in Ads Manager

One of the more operationally significant changes in Q1 was the launch of a native A/B testing tool in Ads Manager, available in beta. Previously, structured testing on Pinterest required external measurement partners or manual budget allocation workarounds. The new tool allows advertisers to run KPI-driven tests comparing Performance+ campaigns directly against standard campaigns from within the platform interface.

For media buyers, this matters because it removes a friction point that has slowed Performance+ adoption among more analytically rigorous advertisers. The Mejuri result cited on the call was produced through exactly this framework - a structured 4-week test with a defined KPI rather than a directional read from aggregate data. Broader rollout of the testing tool is expected to validate or complicate that result across more categories and advertiser types.

Search ranking model: 16,000 user actions over two years

The platform's search infrastructure received a significant update in Q1, separate from the Performance+ product layer. Pinterest updated its proprietary search ranking model to extend the user context window within search by 30-fold. The system now uses up to 16,000 user actions over a two-year period to inform the results shown to each individual user.

This is the same expansion Pinterest previously applied to its Home Feed ranking model, now extended to search. According to Ready, the launch improved search fulfillment by approximately 70 basis points and saves - the action where users pin content to boards - by approximately 390 basis points.

For advertisers with Shopping ads and catalog-based campaigns, this is a structural change in how their products surface in response to user queries. A user's search behavior from 18 months ago now factors into what they see today. That longer context window shifts search from a purely query-intent system toward a taste-based system - which aligns more closely with how Pinterest's Taste Graph functions in other surfaces, but introduces more signal complexity for anyone trying to understand why specific products are appearing or not appearing in search results.

Over 72% of Pinterest's impressions occur across search surfaces, according to the earnings presentation. That figure, combined with the 80 billion-plus monthly searches cited during the call, makes the search ranking model changes directly relevant for any advertiser running shopping or keyword-targeted campaigns.

PinRec: one model across all surfaces

Underneath both search and feed placements is PinRec, Pinterest's proprietary generative retrieval system. Rather than maintaining separate models for each surface - search, home feed, related pins - PinRec is now a single model generating personalized results across all surfaces simultaneously, informed by the full depth of what Pinterest knows about each user's taste and interest.

According to Ready, PinRec was first launched on search and related services in 2025 and was extended in Q1 2026 to serve content globally site-wide. The site-wide expansion improved search fulfillment by approximately 180 basis points and drove a roughly 180-basis-point reduction in CPA and CPC for advertisers - a direct cost implication for performance campaigns.

The practical meaning for campaign managers: the same user signal that determines what organic content a user sees in their home feed now also shapes which ads they are served across search and related surfaces. Audience segmentation and targeting logic that worked under the previous multi-model structure may behave differently as PinRec unifies the signal across surfaces. Advertisers running broad or interest-based targeting in particular should monitor performance data through the transition.

Canvas: AI image generation trained on Pinterest data

Canvas is Pinterest's in-house AI image generation model, built and trained exclusively on Pinterest content rather than general web data. The model already powers the creative optimization layer in Performance+ campaigns, enabling advertisers to dynamically edit backgrounds and transform basic catalog images into lifestyle imagery automatically.

In Q1, Pinterest released a new version of Canvas supporting real-time high-fidelity image editing, with particular improvements in key verticals. According to Ready, the company expects to expand Canvas to enable more creative experiences for both users and advertisers in the coming months.

For creative and production teams, Canvas is relevant because it operates at what Pinterest describes as "an order of magnitude lower cost than leading third-party models" - a consequence of training on Pinterest-specific data rather than requiring general-purpose generative infrastructure. The trade-off is specificity: Canvas produces imagery calibrated to Pinterest's visual aesthetic and quality standards, which may or may not match output requirements for advertisers whose creative assets are designed for other platforms.

Performance+ creative currently uses Canvas to apply background optimization to catalog product images automatically. Advertisers using catalog-based campaigns who have not explicitly opted into Performance+ creative may want to assess what the automated transformations are doing to their product imagery before scaling budgets on those placements.

Bidding directly connected to advertisers' measurement sources of truth

The most technically substantive change in Q1 was a pilot - now being scaled - that connects Pinterest's AI bidding systems directly to advertisers' own measurement sources of truth. This means the bidding algorithm is no longer optimizing solely toward Pinterest-internal signals or standard conversion events. It can instead optimize toward whatever outcome metric the advertiser defines in their own system: customer lifetime value, profit per order, or another metric.

According to Ready, this was in beta in Q4 2025 and rolled out in Q1 2026. In early testing with one advertiser prioritizing lifetime value, the advertiser cited a 15% to 20% improvement in lifetime value ROAS. Pinterest plans to expand this pilot to additional large sophisticated advertisers later in 2026.

This architecture represents a shift in how Pinterest competes for large advertiser budgets. Rather than asking advertisers to trust Pinterest-defined conversion signals, the platform is meeting advertisers at their own measurement layer. The practical implication is that advertisers with mature first-party measurement infrastructure - MTA models, CLV frameworks, custom bidding signals - now have a direct integration path for those signals into Pinterest's bidding stack.

For the broader advertiser base, the Q1 call also flagged planned deeper integrations with key third-party measurement partners later in 2026. This would extend measurement connectivity beyond the largest advertisers' proprietary systems to a wider set of buyers using standard attribution tools.

The clicks-to-revenue gap: an attribution problem, not a supply problem

Ready made a specific claim on the call that advertising practitioners should note. According to him, Pinterest has sent advertisers more than five times as many clicks as it did three years ago - but revenue has not grown at anything close to that rate. His diagnosis is attribution, not supply or demand. "There's a lot more shopping activity that we're driving than what our monetization currently reflects," he said.

The implication is that Pinterest believes it is already delivering significant commercial outcomes for advertisers that are not being credited to the platform in last-click or platform-attributed reporting. The company's response - deeper measurement integrations, proprietary bidding signal connections, and investment in technical sales staff focused on measurement implementation - is aimed at closing that gap over time.

For advertisers using last-click or narrow attribution windows on Pinterest, this framing suggests a potential undervaluation of the platform's contribution. Whether that undervaluation is real will depend on advertiser-side measurement methodology. Advertisers who have run lift studies or view-through attribution on Pinterest campaigns are better positioned to assess this claim than those relying solely on click-based models.

tvScientific integration: CTV audience extension goes live

The tvScientific acquisition closed in Q1 2026, following the December 2025 announcement of the deal. Pinterest has already begun integrating its audience data and Taste Graph signals with tvScientific's algorithms via tvScientific's buying platform, according to Ready.

The early results cited on the call are notable. One early partner - described as a leading home furnishings omnichannel retailer - saw a nearly 190% increase in incremental audience reach and a 159% increase in incremental sales after layering Pinterest's audience data onto its CTV campaigns. Ready also referenced a 27% increase in outcomes and a 65% increase in purchases achieved by applying the Pinterest Taste Graph on top of tvScientific's algorithms, citing this as a demonstration of what the combined data can do beyond Pinterest's owned properties.

These figures are early-stage and come from a single disclosed partner, so they should be treated as indicative rather than representative. The architectural ambition is clear: over time, Pinterest intends to integrate tvScientific capabilities directly into Pinterest Performance+, creating a unified product that spans search, social, and CTV from a single buying interface. That would allow advertisers to extend Pinterest audience segments - defined by commercial intent and visual taste rather than demographic proxies - to streaming TV inventory.

CFO Julia Donnelly confirmed that Q2 2026 infrastructure costs will be partially driven by the full-quarter impact of tvScientific and investment in GPU capacity. The integration costs are therefore already flowing through the platform's expense base, even while the advertiser-facing product is still being built out.

Pinterest Assistant: taking time to get product-market fit right

Pinterest Assistant, the platform's AI shopping assistant, launched in beta in Q4 2025 and remains in a controlled rollout. Ready addressed questions about the assistant directly, emphasizing deliberate pacing. "We want to make sure we're doing this in a way that deepens the relationship between the user and the advertiser," he said on the call, pointing to what he described as false starts by other platforms that had to subsequently pare back more aggressive rollouts.

The model powering the assistant has been materially advanced over the past few months, according to Ready, through both improvements to the underlying open-source model and post-training on Pinterest's proprietary data. The assistant is being built to solve more of the shopping journey while keeping advertisers in the loop rather than disintermediating their customer relationships.

For advertisers, Pinterest Assistant is not yet a live product requiring immediate action. But the design philosophy - keeping brands and retailers as beneficiaries rather than bypassed parties - distinguishes Pinterest's stated approach from some agentic commerce models. If the assistant scales as intended, it would route shopping intent from within the assistant interface to actual product purchase paths connected to advertisers, not to a platform-controlled checkout layer.

Segment dynamics: where growth is coming from

Understanding where Pinterest is growing helps advertisers assess where the platform can scale budgets efficiently. Europe delivered 27% revenue growth in Q1 to $186 million. Rest of World grew 59% to $72 million. U.S. and Canada grew 13% to $750 million. The Europe and Rest of World figures reflect a base that is still significantly undermonetized relative to user numbers - European ARPU was $1.17 in Q1, against $7.12 in U.S. and Canada. The platform has been expanding retail media and programmatic partnerships specifically to accelerate international monetization.

For advertisers with international audiences, the undermonetization of European and Rest of World users means lower competition for ad inventory in those markets relative to the U.S. The trade-off is less developed measurement infrastructure and fewer localized ad product features.

Donnelly noted on the call that international revenue growth is expected to moderate in Q2 due to deliberate leadership and structural changes in the international go-to-market organization. That transition may create short-term pricing and delivery volatility for international campaigns.

Large retail advertisers: tariff headwinds continue

The platform's largest retail advertisers - the segment that has historically driven a disproportionate share of Pinterest revenue - "remained a headwind" in Q1, according to Donnelly, though she noted some improvement later in the quarter driven by ad platform and product improvements. These large retailers are navigating tariff-related margin pressure and pulling back on advertising spend across the industry.

The securities class action filed on March 30, 2026 alleges that Pinterest executives did not adequately disclose the extent of this pressure to investors before a series of stock drops. The litigation is ongoing and Pinterest has not made detailed public statements about it.

For media buyers at agencies serving large retail clients, the tariff dynamic is a live budget planning variable. Pinterest's growing revenue concentration outside large retailers - through mid-market, enterprise, managed SMB, and international advertisers - is a structural shift that may affect auction dynamics and CPMs in certain categories as the composition of active advertisers changes.

Q2 outlook and what changes are still coming

For Q2 2026, Pinterest guided revenue of $1.133 billion to $1.153 billion, representing 14% to 16% growth year over year. Adjusted EBITDA is expected in the range of $256 million to $276 million. The full-year Adjusted EBITDA margin target remains approximately 29%.

Platform changes still in the pipeline for 2026, according to the earnings call, include: broader rollout of the direct measurement integration pilot to additional large advertisers, deeper integrations with third-party measurement partners for the broader advertiser base, continued expansion of Canvas to new creative experiences, further build-out of the tvScientific integration into Performance+, and continued development of Pinterest Assistant.

The launch of Promote a Pin in March 2026 - a simplified self-serve boosting product requiring no advertising experience - represents Pinterest's effort to bring the lower end of the advertiser market onto the platform simultaneously with these infrastructure improvements at the top. The Taste Graph does the targeting automatically, with no manual input required from the advertiser.

Timeline

Summary

Who: Pinterest, Inc. (NYSE: PINS), reporting for the quarter ended March 31, 2026. CEO Bill Ready and CFO Julia Donnelly led the earnings call.

What: Pinterest reported $1.008 billion in Q1 2026 revenue (18% year-over-year growth) alongside a series of ad platform changes: Performance+ at 30% of lower-funnel revenue with adopters growing spend at twice the rate of non-adopters; a new native A/B testing tool in Ads Manager beta; search ranking model updated to use up to 16,000 user actions over two years; PinRec extended site-wide with 180 bps CPA and CPC improvements; Canvas updated with real-time high-fidelity image editing; AI bidding connected directly to advertisers' own measurement sources; and tvScientific integration delivering early CTV results including 159% incremental sales lift for one disclosed partner.

When: The results cover the three months ended March 31, 2026, announced May 5, 2026.

Where: U.S. and Canada generated 74% of revenue. Europe grew 27% and Rest of World grew 59%, both significantly undermonetized relative to user base size. CTV extension through tvScientific is currently prioritized in the U.S. market.

Why: These changes matter for advertising practitioners because Pinterest is simultaneously restructuring how its bidding, measurement, and creative systems work - changes that will affect campaign performance, attribution reporting, and budget efficiency for advertisers already active on the platform. The Q2 2026 guidance implies moderate growth rather than acceleration, making the pace of Performance+ adoption and measurement integration the key variables to watch in the months ahead.

Share this article
The link has been copied!