Pinterest (PINS.US) Q3 conference call: The growth engine is still the engine of international business and is expected to accelerate again in the future
The Zhitong Finance App learned that the photo social networking and shopping platform Pinterest (PINS.US) released financial results for the third quarter of 2026 on October 29 (Thursday) local time. Despite a record high of 661 million global active users (MAU) and double-digit net sales growth, the company's revenue for the quarter and adjusted profit both fell short of Wall Street expectations, and the overall revenue guidance for the holiday shopping season (fourth quarter) fell short of analysts' expectations, much to the disappointment of investors.
According to financial reports, Pinterest's third-quarter revenue was US$1,174 million, up 12% year over year (up 12% in terms of reporting and fixed exchange rate), falling short of analysts' average expectations of US$1.21 billion; adjusted earnings per share (non-GAAP EPS) of US$0.46, falling short of market expectations of US$0.49 million; adjusted EBITDA of US$339 million (profit margin of 28.9%), which was also slightly lower than the forecast of US$348 million.
The trend of the two sets of profit indicators is clearly divided: according to US GAAP (GAAP), the company's net profit for the quarter was only US$22 million, a sharp drop of 77% from US$92 million in the same period last year, and diluted earnings per share were US$0.04 ($0.13 in the same period last year); non-GAAP net profit was US$270 million, up 3% year on year. The main reason for the differentiation was rising costs — total costs and expenses in the third quarter rose from $991 million to $1.14 billion in the same period last year, exceeding revenue. As a result, operating profit fell from 59 million US dollars to 30 million US dollars, and the operating profit margin narrowed from 5.6% to 2.6%.
In terms of cash flow, operating cash flow was US$316 million, down 2% year on year; free cash flow was US$300 million, down 6% year on year, and free cash flow profit margin was 25.6%.
In the third quarter, Pinterest's global monthly active users increased 10% year over year to 661 million, a record high. Among them, users in the US and Canada increased 4% to 107 million; European users increased 7% to 161 million; and users in other regions increased 13% to 394 million.
However, user growth did not translate into revenue at the same time: revenue in the US and Canadian markets grew 14% to US$894 million, showing steady performance; revenue in other regions surged 24% to US$87 million; and in the European market, despite growing users, revenue was only the same as the previous year (US$193 million). Global average revenue per user (ARPU) performance differentiation: ARPU increased 10% to $8.38 in the US and Canada, increased by 9% in other regions, while ARPU in Europe fell 7% to $1.21, dragging down only a slight increase of 1% to $1.80 in global ARPU, further widening the regional liquidity gap.
An additional downside worth noting is a change in Europe's regulatory environment. According to the company's CEO Bill Ready, third-quarter revenue was negatively impacted by new European regulations affecting cross-border sellers.
Since July 1, 2026, the EU has abolished the long-standing duty-free treatment for low-value goods sent from overseas to the EU: low-value packages now have to pay temporary unified tariffs of 3 euros per item; starting November 1, packages from non-EU online stores and direct mail consumers will also face additional processing fees levied by the European Commission. This policy change directly impacted a large number of cross-border direct mail sellers who rely on the Pinterest platform to reach European consumers, which in turn dragged down the platform's advertising monetization in the European market.
What really triggered the sharp drop in stock prices was the fourth quarter guidance. Pinterest expects fourth-quarter revenue of US$1,425 million to US$1,475 million (median value of US$1.45 billion), but the overall range falls short of analysts' expectations of US$1.52 billion — even at the upper limit of the guidance limit, which is only in line with the 12% growth rate in the third quarter; adjusted EBITDA guidance is US$538 million to US$588 million, which is also lower than market expectations of US$605 million. The company says its revenue guidelines have taken into account moderate foreign exchange headwinds.
Pinterest executives stressed during the conference call that Pinterest's users and shopping intentions are still strong. Although short-term growth is being dragged down by European regulations and international sales organization adjustments, the company is opening up new monetization space through AI, performance advertising, international commercialization, and programmatic advertising.
The company's executives pointed out that after EU policy changes, cross-border sellers drastically cut advertising investment. However, cross-border sellers will not permanently withdraw; instead, they will adapt to the new regulatory environment through local warehousing, etc., but this process takes time. Therefore, the company made it clear that cross-border advertising spending is not expected to recover during this year's holiday shopping season, but the situation may gradually improve by the first half of 2027.
Pinterest's commercialization path over the past three years has been: user shopping behavior → performance advertising products → sales organization adjustments → advertisers increase investment → ARPU increase. This model was first successfully tested in the US and Canada (UCAN), and now the company is replicating the same style of play to the international market. Management believes that the US market also experienced regulatory and cross-border trade shocks last year, but Pinterest finally successfully accelerated again. Therefore, although the current situation in Europe will continue for several quarters, this does not mean that the business model has failed. It is also the most important source of confidence for management to re-accelerate international business in the future.
Furthermore, AI is the most frequent long-term growth theme throughout the entire phone call. Management emphasized that Pinterest's strengths are: AI + Pinterest's unique first-party data + visual understanding + user shopping intent. The company uses its own data training and post-training models, and uses a large number of lightweight models and open source weight models adapted to specific tasks. Management even pointed out that for specific application scenarios of Pinterest, an open source model trained after company data can be superior to a closed-source third-party model of a similar scale, while the cost of a single transaction is less than 8% of a comparable closed source model. This means that the company believes that the focus of its AI strategy is not simply to expand the scale of the model, but to use unique data to achieve a higher AI cost performance ratio and directly commercialize services.
If the whole call was reduced to one sentence, the management was actually conveying to the market: Pinterest's current problem is not “whether users have come”, but “how to turn existing users' shopping intentions into revenue faster and more effectively.” As a result, the company is currently doing three things at the same time: First, maintain user growth. Second, turn AI into a commercial tool. Third, replicate the commercialization model already proven by the United States to the international market.
Here are the details of Pinterest's third quarter earnings call.
I. Executive speeches
1. William Ready - CEO and Director
Good afternoon, and thank you all for attending our Q3 2026 earnings conference call.
Before we begin, I'd like to officially welcome James Dibbo. He joined Pinterest earlier this week as our chief financial officer. James is from Amazon and previously recently served as the company's chief financial officer for advertising and media operations. Prior to that, he was the chief financial officer of Amazon's global consumer business.
James has deep experience with one of the largest and fastest-growing advertising platforms in the industry and one of the best shopping experiences in the world. These two aspects are highly relevant to the business we're building, and we're very happy he's joining Pinterest. Later on the conference call, James will share some thoughts and introduce himself. Welcome James.
Before I talk about the specifics of this quarter, I'd like to take a step back and sort out the stage Pinterest is currently in.
Pinterest is a platform for users to find and search for the things they love. We have increasingly become an important destination for visual search and shopping, supported by differentiated AI capabilities and built around commercial intent. Unlike platforms that are built on passive content consumption, Pinterest helps people discover what they love and take action.
In a time of rapid changes in search and consumer expectations, we continue to capture record consumer demand on our platform. This trend covers different regions and generations, and is particularly prominent among Gen Z users.
In our core UCAN market, user engagement is deepening, as the number of searches, created boards, and backlink clicks in the third quarter all grew faster than the number of users.
The number of users and engagement is the clearest measure of a platform's strength, which gives us confidence in long-term development opportunities.
Over the past year, our focus has been on transforming strong user engagement into more lasting commercialization capabilities. We are mainly achieving this goal in two ways: continuing to improve our performance advertising platform and strengthening our market development system, including introducing new leadership across the organization.
Now let's talk about this quarter's performance.
In the third quarter, we recorded a record high number of users for the 13th consecutive quarter and achieved double-digit user growth for the 12th consecutive quarter.
On the revenue side, we have previously indicated that as we enter this quarter, our international business faced two obstacles: European regulatory adjustments to the low value exemption (de minimis) policy applicable to cross-border sellers, and the restructuring of our international market development organization.
The impact of regulatory changes is greater than we expected because policy changes themselves are difficult to predict when they take effect during the quarter.
The contraction in cross-border seller business accounted for about 3 percentage points of the month-on-month change in global revenue growth from the second quarter to the third quarter. The impact of the international market expansion organization restructuring is basically in line with our expectations, and the impact of the contribution is small.
Overall, third-quarter revenue fell below our guideline range. We are not satisfied with this result, and we have a clear plan in place to accelerate our revenue diversification efforts.
Despite this resistance, our adjusted EBITDA remains within the expected range.
This quarter further highlighted the importance of the work currently underway to establish a broader and more sustainable revenue base.
The re-acceleration of UCAN's business provides a clear case for validating the approach we have taken.
Last year, changes in the US regulatory environment affected large retailers and businesses that sell across borders to the US market. Despite this resistance, we continue to improve our effective advertising platform and market expansion execution capabilities to help UCAN's revenue growth rate accelerate again.
Since this year, UCAN's revenue growth has been at a median level of 10 percentage points, which is significantly faster than in 2025.
We are now using the same approach in international markets, and are already beginning to see early signs that these changes are gradually taking effect in international markets.
This convinced us that the actions we are taking are beneficial to the long-term healthy development of the business.
As we said at the beginning of the year, progress is not always linear, and we may face some short-term resistance when making adjustments that are beneficial to long-term development.
As we look ahead, our priorities remain clear.
First, continue to create differentiated visual search, discovery and shopping experiences;
Second, let AI continue to be the core of everything we do;
Third, accelerate commercialization by improving market expansion and measurement systems.
In this context, I'll start with user engagement because it highlights the unique role Pinterest plays for users.
Pinterest has become a search destination, especially a visual search destination.
Users search Pinterest more than 80 billion times a month. The vast majority of these are visual searches, and more than half are commercial.
This is different from the traditional text-based search experience.
People usually come to Pinterest with their own ideas, but often don't know the exact keyword, brand, or product name.
At the heart of this search experience is how we handle intelligent agentic discovery (agentic discovery).
We use AI to proactively provide users with relevant ideas, products, and next steps based on their tastes and intentions.
For many people, shopping isn't a task that requires automation; it's a pleasant activity and a pastime.
Our AI-assisted experiences are designed to enhance the shopping journey users want to participate in, rather than remove users from this journey.
We use AI trained on Pinterest's unique data to understand people's preferences and predict what's likely to interest them next.
But taste is personal, and ultimately the right to choose belongs to the individual.
We believe the best AI experiences should expand people's choices rather than replace them.
That's why Pinterest is committed to preserving all the good parts of the shopping process that people don't want to give up, just as people don't seek to completely automate their entertainment activities.
About 70% of Pinterest's search journey begins with active recommendations from our AI system before users type in any text.
From there, users explore possibilities, visually compare different options, and turn inspiration into action.
This experience continues to drive user growth.
In the third quarter, global monthly active users (MAU) grew 10%.
This trend is particularly evident among Gen Z users, who are our fastest-growing user group.
Currently, Gen Z accounts for 57% of our total users and is our most engaged user group in terms of monthly retention, storage, and usage time.
They also have clear consumer intentions. Two-thirds of Gen Z users said they came to Pinterest to shop.
This intention also extends to commercialization, as the ARPU (average revenue per user) of Gen Z users is already comparable to that of millennial users and UCAN market users.
As Gen Z begins to enter the peak of consumer spending, this situation is conducive to future ARPU growth.
What makes our user engagement particularly valuable is the intention behind it.
As users search, save content, and create boards, they continuously reveal their changing tastes and interests, and this information helps us predict what they want next.
This creates a valuable feedback loop.
The higher the level of engagement, the more relevant and personalized the search experience, giving users more reasons to return.
These unique signals, accumulated over many years, have created an increasingly differentiated experience for users and strong business opportunities for advertisers.
Pinterest Intelligence is the foundation for connecting our proprietary first-party signals, visual comprehension capabilities, and purpose-built AI models.
With this ability, we can provide users with a more personalized discovery experience while bringing better advertising results to advertisers.
We combine proprietary data with compact, application-specific models, and open weighting models that are post-trained in our own environment.
As I said last quarter, for our application scenarios, the open weighting model can outperform the large-scale closed source third-party model because we can use Pinterest's unique data to post-train these models.
At the same time, we achieved a cost per transaction of less than 8% of a comparable closed source proprietary model.
More importantly, we're efficiently expanding these AI capabilities and applying them to highly commercialable scenarios such as Pinterest Assistant.
Pinterest Assistant is now fully open in the US, and we are encouraged by early performance signs.
Although more than half of searches on Pinterest are commercial, about 80% of searches made using Pinterest Assistant are commercial.
Within Assistant, we're building more intelligent agent-like experiences designed to help users complete the discovery journey rather than excluding users from the process, including the Ask model, which is beginning to expand in the US.
We are also further enhancing core usage scenarios through products such as Beauty Guides and Restyle.
Released as a beta in the third quarter, Restyle helps users reimagine a room based on a single photo, such as adding a new piece of furniture, replacing existing items, or exploring a different style.
It combines Pinterest Intelligence's understanding of style with purchasable items in our product catalog to help users see the effects an idea might have in their own space.
Currently, most of Restyle's experience is powered by Pinterest Canvas, an AI image generation model developed in-house.
The same set of intelligent capabilities that improve the user experience is also helping advertisers achieve continuous cumulative performance improvements.
We're continuously improving the way we search and sort ads, using stronger relevant signals and predicting purchase intent.
We're also speeding up the rollout of these improvements so these features can reach advertisers more quickly.
Visual search ads are a new form of funnel advertising currently in testing. They can help performance advertisers reach some of the most intentional moments of users on Pinterest, including search results pages and pin close-up pages.
This ad format is designed to drive measurable backlink action and provide advertisers with a new way to turn visual discoveries into conversions.
Advertisers can use existing product catalogs for keyword targeting, making it easier to adopt this format.
We are also extending the same effect direction to application promotion.
App promotion is our new goal for app installs and is now open for testing to eligible advertisers around the world.
Pinterest hasn't been a significant player in mobile app advertising in the past, and this has opened up a significant new budget pool for us.
With the launch of these products, we're also making Pinterest easier to use, optimize, and scale up for advertisers.
We've always made it clear that Pinterest Performance+, an AI-driven advertising solution, will be a multi-year product and adoption cycle.
Currently, advertisers can add the keywords and interests that matter most to them, and track the results for each item separately.
With priority product features, they can also highlight the product categories they value most.
The end result is greater control and transparency while automating.
Big advertisers have always wanted both, and they're driving continued adoption of Pinterest Performance+.
We're also advancing measurement and bidding capabilities to help advertisers see the full value Pinterest creates.
In the third quarter, more and more of our mature advertisers directly integrated their internal measurement systems with our AI bidding tools.
These integrations allow us to optimize for the results each advertiser values most.
For example, in an early test with a leading e-commerce platform, the advertiser saw a 77% increase in return on advertising investment (ROAS) for new customers through its internal measurement system.
We are combining these platform improvements with more stringent market expansion methods to focus on driving performance ad sales.
Looking back, our performance advertising business has only been developing for about three years. The business initially started in the UCAN market.
We built a funnel business model around shopping and adopted a conscious order of promotion: first make Pinterest a place for users to shop and take action.
We then set up a performance advertising platform to enable advertisers to take advantage of this user intent and allow sales and market development teams to adjust around this.
Now we're using the same approach in international markets.
In terms of market expansion, we are advancing in three areas: people, organizational structure, and operational discipline.
First, there are people.
We've empowered our sales leadership team to align with our shift to effective sales because this sales approach requires different skills than upper-level sales funnel brand advertising.
In international markets, we've only recently begun this process, so more adjustments are needed in multiple markets.
Overall, the hardest part of staff adjustments is over, but many of our new leaders are still in the run-in phase.
The second is the organizational structure.
We're optimizing customer coverage around maximum opportunities and getting technical and sales teams to work more closely together.
This allows advertisers to more consistently get the performance, measurement, and product expertise they need to grow their Pinterest business.
The third is operational discipline.
We're establishing a clear sales staff responsibility system through more consistent training, incentives, and sales processes.
In the UCAN marketplace, this helps us package key commercial moments more effectively and more directly connect Pinterest's product performance to advertiser results.
In international markets, we are applying the same discipline to every market.
These actions are aimed at improving customer reach, sales force productivity, and our ability to turn strong advertiser results into more lasting revenue growth.
At the end of the day, our goal is to create a more resilient revenue base that reflects both the strength of users and engagement, and the results we bring to advertisers.
In summary, I am confident in this work because we have a very strong foundation.
Our users are growing and remain highly engaged.
Our platform is delivering results for advertisers, and we are also strengthening our market expansion engine.
Our current work will enable Pinterest to re-accelerate growth over time and restore the medium to high 10 percentage point revenue growth achieved in previous history.
Before I talk about the results and outlook, I'd like to ask James to talk about why he joined Pinterest.
2. James Dibbo — Chief Financial Officer
Thanks Bill.
I'm excited to join Pinterest on Monday. Since I've only been with the company for a few days, today's statement will be rather brief.
I joined Pinterest because I really admire everything Bill and the team have built, but the main reason is that we still have huge upside opportunities in the future.
I first came into Pinterest in 2014, and I've been using Pinterest ever since, including for moving and decorating, and planning my dream Jeep Wrangler modification plan.
Just this past weekend, I had fun searching for all kinds of outfit ideas to prepare for my first day as Pinterest's CFO.
I'm looking forward to learning more about the analyst and investor community and working with Bill, the board, and the entire team to realize Pinterest's full potential as the world's leading visual search, shopping, and advertising platform.
Pinterest is a place where a growing number of users from all over the world find inspiration and discover ideas, products, and experiences that bring joy to offline life in a safe online space.
At a time when more people are actively trying to reduce social media usage, as a father of two daughters, I think Pinterest's focus on positive and inspiring content is particularly timely and important, and also a huge business opportunity.
The AI-driven innovations the team continues to develop are helping more and more brands reach consumers with relevant, useful, and actionable ads at the right time.
These ads complement the user's Pinterest experience, and if successful, they will form a flywheel of ever-accelerating growth: increased advertiser value, deeper user engagement, and increased shareholder value.
I'm looking forward to the next trip and to talking with you on future earnings calls.
Thank you all, and now return the phone to Bill.
3. William Ready
Thank you, James.
I will now discuss our financial results for the third quarter of 2026 and update our outlook for the fourth quarter of 2026.
With the exception of revenue, all financial measures will be discussed on a non-GAAP basis unless otherwise stated; unless otherwise stated, all comparisons are year-over-year comparisons.
By the end of the quarter, we had 661 million monthly active global users, or MAU, up 10% year over year, once again reaching a record high.
We continue to grow our users in all regions.
In the third quarter, the US and Canada had 107 million MAU, an increase of 4% year over year.
The European region had 161 million MAU, an increase of 7% over the previous year.
In other markets around the world, we have 394 million MAU, an increase of 13% year over year.
Let's move on to revenue.
In the third quarter, our global revenue was $1,174 billion, up 12% in terms of reporting and fixed exchange rates.
We consider the target advertising business to perform strongly, especially performance-oriented CPC advertisers; at the same time, brand awareness also contributed to the performance of targeted advertising.
From an industry perspective, retail continues to perform strongly, while some smaller but fastest-growing emerging industries on the platform are also performing well, including financial services, healthcare, and tourism.
Let's take a look at the performance of each region in the third quarter.
Revenue in the US and Canada was $894 million, up 14% year over year.
Growth is mainly from retail, financial services, and healthcare.
European revenue was US$193 million, roughly the same as reported, and increased by 1% on a fixed exchange rate scale.
The contraction of cross-border seller business dragged Europe's year-on-year growth by about 11 percentage points.
Revenue for the rest of the world was $87 million, up 24% at the reporting level and 21% at a fixed exchange rate.
I would like to clearly explain the main reason why global revenue growth slowed from 18% month-on-month in the second quarter to 12% in the third quarter.
About 3 percent of this is due to the contraction of business by cross-border sellers in response to regulatory changes in the EU's low-value tax exemption policy.
About 2 percent comes from the timing of Prime Day and World Cup-related spending, which we mentioned last quarter.
The remaining 1 percentage point comes from foreign exchange factors. Foreign exchange contributed 1 percentage point to growth in the second quarter, while the impact in the third quarter was neutral.
Overall ad impressions increased 17% in the third quarter.
As ad impressions grew faster than revenue growth, advertising prices fell 4% in the third quarter.
Let's talk about fees below.
Revenue costs for the third quarter were US$250 million, up 21% year over year and 2% from the second quarter, mainly driven by increased infrastructure spending associated with increased users and engagement, investment in GPU capacity, and TVScientific.
Our non-GAAP operating expenses were $591 million, up 9% year over year.
The increase is mainly due to sales and marketing expenses, including higher marketing expenses and an increase in the number of sales staff, as well as R&D expenses to support AI and product plans.
By the end of the third quarter, we had 5,192 employees, a slight year-on-year decrease, including employees who joined Pinterest through the acquisition of TVScientific.
Excluding this acquisition, the number of employees fell 4% year over year in the third quarter.
We will continue to invest in professionals in areas where we can see a clear and attractive return on investment.
But at the same time, one of our continuing focus is finding the best combination of talent, infrastructure, and AI-enabled productivity tools to drive lasting growth and increase profit margins over time.
In the third quarter, we achieved an adjusted EBITDA of US$339 million, and the adjusted EBITDA margin was 29%, down 30 basis points from the third quarter of last year.
Our free cash flow for the third quarter also reached $300 million.
Over the past 12 months, we have generated nearly $1.26 billion in free cash flow, and our free cash flow conversion rate has reached 90%.
As of the end of the quarter, we had $1.5 billion in cash, cash equivalents and marketable securities.
Now let's talk about fourth quarter results guidance.
We expect fourth-quarter revenue to be between $1,425 million and $1,475 million, up 8% to 12% year over year.
Based on the current spot exchange rate, we expect FX to form moderate resistance in the fourth quarter.
Our outlook assumes that the factors affecting third-quarter revenue growth will continue in the fourth quarter.
First, we expect the UCAN region, our largest market, to continue to experience strong revenue growth.
However, we expect that the contraction in cross-border seller business will continue to drag down international revenue growth.
We are already beginning to see early signs that changes in international market expansion are taking effect, but the relevant revenue will take time to gradually accumulate, and it is expected that the fourth quarter will not fully offset this pressure.
To remain cautious, we have extended our fourth-quarter revenue guidance range to reflect the broader results we believe are likely to occur this quarter.
We are confident in the fundamental drivers of our business and have made significant progress to make our operations simpler and focus our investments in the areas that matter most, giving us the flexibility to respond to a range of different scenarios.
For the fourth quarter, we expect adjusted EBITDA to be between $538 million and $588 million.
We expect non-GAAP revenue costs for the fourth quarter of 2026 to be ten percentage points lower than the month-on-month increase in the third quarter of 2026, partly due to our investment in GPU capacity.
As a reminder, non-GAAP revenue costs for the third quarter of 2026 benefited from the accelerated recognition of certain contractual benefits in our recently signed multi-year infrastructure agreement.
Among non-GAAP operating expenses in the fourth quarter, the main areas of year-over-year investment growth will be sales and marketing, and R&D to support AI and product plans.
We continue to expect an adjusted EBITDA margin of around 30% for the full year of 2026.
Although recent revenue growth has fallen short of our target, I'm proud of the experience we continue to provide to our users.
I am confident in our team, strategy, and actions we are taking, and I believe these measures will propel continued growth to re-accelerate over time.
II. Q&A session
1. Oppenheimer analyst Jason Helfstein
Bill, I think you've explained the resistance you faced in the third quarter very clearly. I'm interested in what you mentioned “Internally is already seeing signs of a possible acceleration in international business.” Obviously, we already have fourth-quarter results guidance. But I'd like to know more, when do you think these initiatives will begin to reverse international revenue trends? Also, if there is any other information that can help us understand the situation in the third quarter and how this situation will continue into the fourth quarter and even next year, please let us know.
William Ready
Thanks for the question.
I'll go a little further on what happened in the third quarter before going into some details about our outlook for the fourth quarter.
In the third quarter, various favorable and negative factors in the business were consistent with the situation described in our last earnings call. That is, the UCAN region continued to achieve strong growth, while international business was affected by two resistances.
One is a change in European regulatory policy, and the other is a restructuring of our own market development organization.
However, the impact of changes in European regulatory policies exceeded our forecasts, causing growth to slow by 3 percentage points during the quarter.
Let me elaborate more on these factors.
We continue to achieve strong user engagement growth globally and in the UCAN market.
The number of users increased by 10% to 661 million.
User engagement in the UCAN region is deepening, and the number of searches, number of boards created, and backlink hits are growing faster than the number of users.
UCAN's revenue has increased 15% since this year, which is a significant acceleration from 2025.
Globally, revenue growth slowed from 18% month-on-month in the second quarter to 12% in the third quarter.
Of these, 3 percentage points came from the factors we mentioned in our last earnings call, including the World Cup, the time change in Prime Day from the third quarter to the second quarter, and 1 percentage point contributed by foreign exchange factors.
Another 3 percentage points came from the cancellation of Europe's low-value tax exemption policy, which caused the cross-border seller business to shrink more than we expected.
Such regulatory changes themselves are difficult to predict, especially when policies are in effect during the quarter.
We have also been affected by some short-term disruptions caused by the restructuring of international market development organizations, but this impact is largely in line with our expectations.
It should also be added that in recent months, in the process of transitioning to a more performance-oriented sales model, we have made a number of leadership adjustments.
In fact, we are very encouraged by the new talent joining the company. The most difficult phase of transformation is almost over, and we are beginning to see early signs that these changes are taking effect.
For example, the growth rate of our core European corporate business did not include the cross-border seller business, which halved in the third quarter compared to the second quarter.
However, since entering the fourth quarter, we have seen this growth rate accelerate again and exceed the level of the second quarter.
As a result, while we and advertisers dealt with major regulatory changes, policy implementation throughout the quarter, and ultimately led to revenue falling short of guidelines, we saw deviations from forecasts become more apparent later in the quarter.
However, after entering the fourth quarter, the situation has stabilized.
It is emphasized once again that we are seeing positive early signs of the restructuring of the international sales organization.
As for the fourth quarter, based on the results so far, the overall growth in the fourth quarter was basically the same as that of the third quarter.
We continue to see strong global user engagement and continued strong growth in UCAN's revenue; at the same time, the cross-border seller business continues to put pressure on international market growth, but this pressure has stabilized.
As mentioned earlier, we have expanded our revenue guidance range to reflect a wider range of potential results, while also hoping to provide a prudent benchmark, and we have high confidence that we can reach or exceed this level.
2. John Blackledge, analyst at TD Cowen
Very good. I think you've already answered part of it, but I don't know if you can further discuss the impact of regulatory changes and low-value tax exemption policy changes on cross-border sellers. Also, continuing with Jason's question, how do you see the European market resuming growth, considering that we are about to enter next year? It would be really helpful if you could provide some information.
William Ready
Thanks for the question.
As I mentioned, these types of regulatory changes are difficult to predict in the short term because advertisers need to adapt to new realities.
As I mentioned, this factor changed the overall global revenue growth rate by 3 percentage points from month to month.
We expect this factor to continue to drag down international market growth in the fourth quarter, but the impact has stabilized so far.
But I think it's also important to review what happened to these advertisers last year.
When US tariffs came into effect, many advertisers redirected their spending to Europe and Latin America.
Today, both the US and Europe have experienced trade barriers, and there are fewer major markets to transfer.
So in the short term, we're seeing these advertisers cutting spending globally to protect profit margins.
But we expect these sellers will eventually adapt.
For example, some retailers have already mentioned that they plan to use local warehouses to deal with new regulations.
But these adjustments will take some time, so it's hard for us to predict them.
Meanwhile, our new sales leadership team is gradually getting into shape.
They're helping us bring in more advertisers in more categories to diversify our revenue over time.
So while this doesn't completely offset the resistance we're facing, we're still quite encouraged by it.
Although we currently have no guidance for next year, and we don't think that cross-border advertising spending will recover during the holiday season, we think that looking ahead to the first half of next year, we will have more time to complete similar adjustments for these advertisers, and we will continue to diversify our revenue.
3. Baird analyst Colin Sebastian
I'd also like to welcome James on board. Bill, thank you for talking about smart agent-related initiatives. What do you think of the relationship between Pinterest Assistant and the app's core discovery features? Also, I'd like to better understand your current understanding of the role agents or assistants play in helping users move from inspiration to purchase.
William Ready
Thanks for the question.
First of all, I would like to say that at a time when we have reached a record high number of users for the 13th consecutive quarter, we have continued to achieve strong user and engagement growth even though various AI-driven experiences and intelligent agent experiences have appeared in the market.
At its core, we're providing an AI-driven shopping experience, which we believe really satisfies the way users want it.
I also mentioned a few things in my prepared statement.
From the beginning, we've focused on the part of the shopping process that users really want to participate in.
You've often heard me say that the first 25 years of e-commerce development solved the problem of “buying” to some extent, but in some ways they killed “shopping.”
The difference between the two is that “shopping” is a journey users want to take part in, including the joy of discovery, figuring out what style they want, and constantly improving their own taste.
So when we designed Pinterest Assistant, our first consideration was to incorporate it into the overall experience.
It acts as a co-driver, helping users move from inspiration to comparison and research, while keeping users in control at all times.
Pinterest Assistant is our AI conversation layer, and we don't treat it as a standalone experience.
It's integrated into the overall user experience, and every PIN on Pinterest itself is already powered by AI.
But it does bring large-scale language model capabilities to our visual, personalized experiences, enabling users to research, compare, and make decisions without leaving Pinterest.
The experience is now fully open in the US, and we're very encouraged by how many users have participated in it.
One of the most encouraging signs we've seen is that around 80% of Assistant searches are commercial, compared to about half of other searches on Pinterest.
This shows that we're matching AI to highly commercializable use cases.
We are very satisfied with the match between this AI application and commercialization, and our ability to commercialize it.
We do over 80 billion searches every month.
As for smart agents, currently about 70% of searches on Pinterest start with active recommendations.
This gives you a sense of the extent to which we've incorporated proactive recommendations into our core experience, proactively providing recommendations to users before they type or ask for anything.
This reflects the promise of intelligent agents to some extent, but it is also achieved in a way suitable for large-scale user adoption, allowing ordinary users to participate in these experiences according to their own habits.
As a result, Assistant makes recommendations easier to understand and helps users where they are.
Looking ahead, we do see opportunities to further expand the smart agent experience, such as the Ask model I mentioned earlier.
Currently available to millions of users in the US, the feature can help users complete multi-step journeys, such as planning trips.
For example, when planning a trip, you can ask how to plan a trip. The system will recommend places to stay, possible activities to attend, restaurants you can go to, and clothes you can prepare based on your style and taste.
All of these features are helping us to further deepen the intelligent agent experience.
But again, we keep users in control: they can get great recommendations, but ultimately it's up to the users to decide.
We're just getting started on this.
We will continue to focus on expanding enforceability and commercialization, continuously testing, innovating, and adding ancillary features.
But once again, we are very satisfied with how to combine these AI-driven capabilities with a good business experience, which will facilitate future commercialization.
We're also very happy with the way these experiences are designed for large users, to help users where they are and help them make the decisions they want to make while still leaving them ultimately to decide for themselves.
We think this is very important.
People don't want to lose control of their decisions.
They may want help making decisions, but they don't want to lose the power to make decisions.
Therefore, we believe this is at the core of driving strong user and engagement growth.
4. Justin Patterson, KeyBanc Analyst
I'd like you to talk more about the decision to expand programmatic partnerships. How do you see the impact of this move on advertiser coverage and bid density? Also, since you've had some similar collaborations in the past, can you talk about your initial experience and how it's progressing so far?
William Ready
Thanks for the question.
Yes, we're expanding our range of programmatic partners, which is one of the many ways we expand Pinterest's demand sources and help us reach advertisers where they are.
Until now, our goal has been to supplement first-party businesses with more sources of demand, fill gaps in our advertising bids, and take advantage of the commercial intent we've seen on the platform.
However, we also see a very meaningful opportunity to make it easier for customers to buy ads the way they want.
As we see more advertisers programmatically buying ads, we want to get closer to them.
As we announced this week, our partners currently cover approximately two-thirds of the DSP market, including Google DV360, Amazon DVA+, and Yahoo! DSP and Magnite are our supply-side partners.
Looking ahead, we think there will be more similar opportunities.
As we've shared over the past few quarters, we're also seeing an important opportunity to tap into our audience outside of the Pinterest platform.
We began in this direction with our acquisition of TVScientific earlier this year, which extended our audience reach to connected television (CTV).
As these platforms continue to evolve, we can leverage our unique audience data in more privacy-preserving ways by partnering with these DSPs.
This opens up an opportunity for us to programmatically bring in more demand while continuing to highlight the value of our unique signal and Pinterest's highly unique, commercial audience.
Finally, I'd like to remind you that, as I've shared before about these collaborations, this kind of requirement takes time to debug and climb, as did programmatic collaborations in the past.
So we think this is an important step to make it easier for more partners to bring more demand into the platform.
However, it is emphasized once again that it is expected that these businesses will take time to gradually climb.
5. Mizuho analyst Lloyd Walmsley
I'd like to ask about the new ad format for visual search that you recently announced. There seems to be some very positive data in the information you've posted over the past month or so. I remember you saying that the vast majority of searches are visual, and more than half of them are commercial. But can you help us understand the potential significance of this business? Is it comparable to other big services? How big is the potential incremental space for this new ad format?
William Ready
Thanks for the question, Lloyd.
Yes, visual search is a new way to search online.
We are a leader in visual search, with over 80 billion searches per month, the vast majority of which are visual searches.
We're launching visual search ads to allow advertisers to directly participate in this experience for the first time.
We think this is a huge and significant opportunity.
These visual searches on the platform have high commercial intentions, and more than half of them have commercial attributes.
Now, we're really opening up a way for advertisers to interact with users at the very beginning of these interactions.
We used to focus on creating these experiences in a natural way first, and didn't actually get ads into visual search experiences.
Now, we're opening up this field to advertisers, so advertisers can get more prominent and advanced ad positions.
As a result, visual search ads will appear at the top of visual search results, and will also appear on the Pin close-up page, which is where the user is viewing the product they want to buy.
This unlocks opportunities for advertisers to interact with users.
They can now interact with users in ways other platforms can't, especially considering Pinterest's highly visual interactions.
This is a very important user need we're addressing, and it's different from the experience offered by traditional search or chatbots.
One of the biggest problems traditional search and chatbots face is that users often don't have enough language to accurately express what they're looking for.
They probably just have a vague idea.
So that's why I mentioned before that over 70% of our searches start with unsolicited recommendations.
When you combine our proactive recommendations with the fact that users don't have to ask, this is in part the promise of an intelligent agent experience — helping users and acting on their behalf.
Then, in combination with users being able to further participate, visually refine their needs, and get high-quality product displays and advertisements relevant to them at the top of the results.
We think this is a very engaging experience for both users and advertisers.
The feature debuted on Pinterest Presents in September and is currently in beta.
So we're really excited about it.
We're learning with advertisers and look forward to sharing more relevant information as the feature evolves.
6. LightShed Partners analyst Rich Greenfield
I have two questions. First, I'd like to follow up on the DSP collaboration. Why now? Obviously, you could have carried out these collaborations last quarter or even last year. Bill, what was the catalyst that brought these collaborations to fruition earlier this week? Also, I remember that last quarter, investors were very excited about the increase in US revenue, because the revenue growth rate in this key market accelerated from about ten percent to ten percentage points higher. And this quarter, your growth rate has slowed to about 14%. Are there any particular factors driving the previous acceleration and this quarter's slowdown? Or is this higher level of growth sustainable as we consider future developments?
William Ready
Thanks Rich, and thanks for the question.
I'll answer the first question and then let Andrew answer the second.
As for DSP collaboration, you've been listening to us talk about this for a while.
We believe there are plenty of opportunities to commercialize and apply our audience to other scenarios.
We first achieved this goal by introducing third-party requirements into our platform, then expanded this capability beyond the platform through the acquisition of TVScientific.
Part of what drove us further was our experience when we first introduced demand to the platform and the experience we gained in commercializing audiences in new scenarios such as connected TV.
At the same time, the entire industry has also advanced in privacy protection technology, which really helps us think about how to collaborate with other partners, reach out to advertisers where they are, bring Pinterest's unique audience and our outstanding personalization capabilities to more scenarios, or collaborate with data from other partners.
As for why now, part of the reason is that we have been gradually advancing through testing and learning, and advances in privacy protection technology have convinced us that we can conduct these businesses in a way that not only provides a better experience for users and advertisers, but also protects privacy.
At the same time, we are also able to ensure that when we carry out these collaborations, we get the value created by our audience.
Let's leave the second question to Andrew.
Andrew Somberg — Vice President of Investor Relations and Finance
Thanks Rich.
UCAN grew 14% in the third quarter, which represents a strong re-acceleration compared to the level at the end of 2025.
As mentioned in our last conference call, the second quarter included some special one-off benefits, including additional expenses from the World Cup and the transfer of Prime Day to the second quarter.
As a result, in the third quarter, we saw some weakness in the food, drink and alcohol industry, which put pressure on the UK market by about 1 percent month-on-month.
Meanwhile, as we mentioned, we're successfully expanding our advertiser base in different categories to help offset some of the pressure in the food and beverage sector.
As a result, in addition to retail, we've also seen the healthcare industry perform well, including the growth of GLP-1 advertisers, which helped partially offset the pressure on the food and beverage industry.
As a result, looking ahead to the fourth quarter, we are generally satisfied with the UCAN market.
But these are some of the pros and cons I would point out to Rich.
William Ready
Yes.
I want to emphasize one last point. The re-acceleration of UCAN's business and our very strong user growth and engagement are all real highlights in the business.
Last year, we dealt with regulatory changes in the US, similar to what we are currently experiencing in Europe.
But we've clearly achieved a fresh acceleration from last year.
It's not just our own situation; you've also heard other companies talk about the World Cup, Prime Day, and how these factors changed the way they do business in the second and third quarters.
This is largely in line with what you've seen in the market.
However, if you compare the UCAN situation in the same period last year, you can still see a renewed acceleration very clearly.
We are happy with that.
As you've heard from our performance guidance, we expect the strong momentum in the UCAN market to continue into the fourth quarter.
7. Goldman Sachs analyst Eric Sheridan
I have two quick questions. First, based on the user growth you've achieved, I'd like to further ask about the topics we've been discussing during today's conference call. How should investors consider the duration of this opportunity compared to the already achieved user growth and the ARPU or commercialization gap you still have a chance to close? We want to understand how these two might relate to each other over the next 12 to 24 months. Also, I think your previous answers were mostly about Pinterest Assistant. If we expand our discussion to horizontal intelligent agents, how do you see the opportunities they bring and some of the friction points that need to be addressed?
William Ready
Thanks for the question, Eric.
Regarding ARPU growth, I think you can take a look at our US business, which is the UCAN market where we first started our shopping business.
As a reminder, we only launched our first specially designed click-based advertising product to advertisers about three years ago.
As a result, we're still in the early stages of our commercialization journey, but we've made very good progress.
As we reinvent Pinterest as an AI-powered shopping platform, we're clearly attracting users here, and you've seen ARPU improvements in the UCAN marketplace.
Looking ahead to the international market, we have also made good progress.
However, if we compare the ratio of ARPU in the European or international market to the ARPU of the US business, we can see that there is still a clear gap between us.
The user behavior we see, that is, shopping-type user behavior, is a global phenomenon.
So for us, user growth and engagement driven by shopping behavior is a global phenomenon.
In European and international markets, we initially pioneered advertising opportunities in the largest domestic markets, and are now expanding this model to these markets.
So while we're still in the early stages, we're implementing the same business model and are already seeing it start to pay off.
Changes in European regulatory policies have caused us to experience a small “speed bump”, but this has not affected the growth and participation of users in shopping.
So at the end of the day, if you have user engagement when it comes to shopping, advertisers will revolve around that engagement.
We are making better use of this level of engagement by advertisers through the realignment of advertising products and market expansion systems.
As I mentioned, we are seeing some early positive signs.
So we're excited about this development.
But at the same time, we're also very excited because there's plenty of upside here.
When you look at how low our ARPU level is and its relative position compared to the US market, and consider similar ratios on other platforms, you'll find that we have a lot of room to rise by increasing ARPU in the international market.
Looking ahead to next year, we see that the ongoing international market expansion adjustments are beginning to bear fruit.
I've shared a few examples.
What we've seen in the US is that it may take a few quarters for advertisers to adjust to tariff changes.
We can't accurately predict how changes to Europe's low-value tax exemption policy will eventually unfold, but this is likely to be an event that will continue for multiple quarters.
However, just as we have achieved a new acceleration in the US, we think the international market can also achieve a new acceleration.
The most important factor that gives us confidence is that we see that core user shopping behavior still exists.
So looking ahead to next year — and again, we won't be providing performance guidance beyond the quarter — but looking at next year, we expect to be able to accelerate again, just like we did in the US.
Regarding your second question, you mentioned horizontal smart agents.
What I think is very interesting is that if you look back over the past 20 years of search development, you can see that although there are a few winners in the horizontal search field, there are also many winners in the more vertical search experience.
For example, how many product searches will start on Amazon rather than on a horizontal search platform?
And how many travel searches start with Booking or Expedia rather than a horizontal search platform?
As we look to the future world of smart agents, we think there will be a wide range of horizontal smart agents, but there will also be smart agent winners focused on specific vertical sectors.
We are very focused on ensuring a great shopping experience and that this experience helps users continue to drive relevant activities.
Therefore, we are very pleased with the progress so far.
We've been designing in such a way that we can deliver services where users are, thereby driving mass user adoption.
We're providing an intelligent agent experience where users feel like it's helping and assisting themselves rather than taking decisions away from them.
I've shared some relevant data.
I think that's why after many rounds of chatbot improvements and now other horizontal smart agents have appeared, we haven't seen UCAN user engagement affected.
Because these tools tend to be more horizontal, users come to Pinterest for the specific purpose of shopping.
This is an area where we can truly deliver highly personalized experiences, and users feel they are still in control of the entire experience.
Simply put, users are willing to automate tedious tasks.
But they don't want to automate their happiness.
They don't want to automate their entertainment and pastime.
We did some research on this.
A recent study conducted by YouGov on behalf of Pinterest found that more than 80% of Americans said that if AI completed their shopping for them, they would miss certain things in shopping, including discovering things they weren't looking for and figuring out what they liked.
That's why Pinterest has always been committed to preserving all the good parts of the shopping process that people don't want to give up, just as people don't seek to automate entertainment activities.
As such, this has always been at the core of our continued deepening user engagement.
As I mentioned earlier, in the US, we continue to see increased user engagement, particularly among Gen Z, the digital native user group.
Even as other tools keep popping up, we're seeing continued deepening user engagement.
One important reason is that we're in the midst of an expansion of search.
I think we're carving out a unique space for ourselves, especially in the field of shopping.
We do this in a highly personalized way, while striking the right balance between users maintaining autonomy, participating in decisions, and completely entrusting decisions to others.
As the study I just mentioned shows, users don't necessarily want to give up their decision-making power.
They want help in the decision-making process, but they still want to participate in the decision-making process.
As a result, this trend will continue to develop. We have more work to do, but we are very pleased with the progress we have made so far.
8. Shweta Khajuria, analyst at Wolfe Research
I'd like to ask a question about bid optimization. The AI-driven bid optimization feature you deployed at the end of the first quarter and during the second quarter received very positive feedback from major advertisers. Can you tell us more about how this feature has progressed this quarter? How should we consider the potential tailwind effects of this feature between now and the end of the year and beyond?
William Ready
Thanks for the question, Shweta.
As you pointed out, we're aligning the AI bidding system with the advertisers' real measurement results source.
This allows us to achieve the highly specific results our advertisers care about, and we've seen strong results.
We've made significant progress in providing our users with highly personalized recommendations to their shopping needs.
But every advertiser makes a different judgment on the value of different events, and they want to see different results.
When advertisers provide us with more information about the value of different events — for example, whether they want to focus on selling specific products, whether they want to acquire new customers, or drive deep engagement with existing customers — they can tell us what they want to achieve.
Just as our AI can provide highly specific recommendations to consumers, now we can also provide highly specific results for advertisers.
We are therefore very pleased with the progress made in this regard.
So far, we've mainly focused on mature advertisers with proprietary measurement systems.
As we said earlier this year, these advertisers are being more active in using our products as a result.
In the third quarter, as planned, we expanded these integrations to a few other mature advertisers.
For example, in an early test with a leading e-commerce platform, the advertiser saw a 77% increase in ROAS for new customers through its internal measurement system.
We are happy with that.
We will continue to expand this range.
This is also a very important moment as the entire industry is gradually moving beyond last-click attribution.
This is ultimately beneficial to platforms like Pinterest, but different advertisers are not moving at the same pace.
We're also continuing to learn from these direct integrations and further expand this capability to a wider group of advertisers by partnering with third party measurement service providers.
For more mature advertisers, we use direct integration.
We're using the experience we've gained to work more with third-party measurement service providers to reach a wider group of advertisers.
These adoption curves will naturally take time, but we're very encouraged by what can be achieved here.
Once again, we've gone a long way in terms of consumer and user engagement.
But the technology we use for consumers to drive highly personalized results can now also help us match highly specific results for advertisers, as advertisers are providing more specific attribution information for different events.
Therefore, these things will take time to fully reflect, but we are very satisfied with their significance for the company's long-term development.
Finally, I would like to thank everyone who participated in our conference call today and asked questions.
Before I finish, I'd like to thank Julia Donnelly for her contributions to Pinterest.
Finally, I'd like to thank the team at Pinterest, our advertising partners, and everyone who came to Pinterest to find inspiration and take action.