Pinterest, Inc. (NYSE:PINS) Q2 2023 Earnings Conference Call August 1, 2023 4:30 PM ET

Company Participants

Neil Doshi – Head-Investor Relations

Bill Ready – Chief Executive Officer

Julia Donnelly - Chief Financial Officer

Conference Call Participants

Eric Sheridan - Goldman Sachs

Ross Sandler - Barclays

Lloyd Walmsley - UBS

Mark Mahaney - Evercore

Rich Greenfield - LightShed Partners

Doug Anmuth - JPMorgan

Operator

Good afternoon, and thank you for attending the Pinterest Second Quarter 2023 Earnings Conference Call. My name is Elisa, and I will be your moderator for today's call. All lines will be muted during the presentation portion of the call with an opportunity for questions and answers at the end. [Operator Instructions]

I would now like to pass the conference over to your host, Neil Doshi, Head of Investor Relations. You may proceed.

Neil Doshi

Good afternoon, and thank you for joining us. Welcome to Pinterest's earnings call for the second quarter ended June 30, 2023. I'm Neil Doshi, Head of Investor Relations for Pinterest. Joining me today on the call are Bill Ready, Pinterest's CEO; and Julia Donnelly, our CFO.

Now I'll cover the safe harbor. Some of the statements that we make today regarding our performance, operations and outlook, may be considered forward-looking, and such statements involve a number of risks and uncertainties that could cause actual results to differ materially. In addition, our results, trends and outlook for Q3 2023 and beyond are preliminary and are not an indication of future performance. We are making these forward-looking statements based on information available to us as of today and we disclaim any duty to update them later unless required by law. For more information, please refer to the risk factors discussed in our most recent Forms 10-Q or 10-K filed with the SEC and available on our Investor Relations website.

During this call, we will present both GAAP and non-GAAP financial measures. A reconciliation of non-GAAP to GAAP measures is included in today's earnings press release and presentation, which are distributed and available to the public through our Investor Relations website located at investor.pinterestinc.com. Lastly, all growth rates discussed in today's prepared remarks should be considered year-over-year unless otherwise specified.

And now I'll turn the call over to Bill.

Bill Ready

Thanks, Neil, and thank you all for joining our second quarter 2023 earnings call. Q2 was a solid quarter marked by strong growth across our business. We're seeing ongoing momentum, giving us continued confidence that our strategy is yielding results with more potential as we look ahead. We ended the quarter with 465 million MAUs, up 8%. Our total revenue was $708 million, growing 6% or 7% on a constant currency basis. We stay disciplined with our expenses and delivered adjusted EBITDA of $107 million and an adjusted EBITDA margin of 15%.

Before we dive into the key highlights from Q2, I'd like to share more about what we've accomplished over the last year since I joined and why I'm excited for the opportunities ahead. During this time, we've become laser-focused on our strategy of building upon our core differentiators and visual discovery to help our users go from inspiration to action.

With that focus on our strategic priorities, we've successfully returned to strong user and engagement growth while delivering consistent year-on-year revenue growth and returning to margin expansion despite the downturn in the advertising environment.

On the engagement side, we've utilized next-generation AI technologies to service more relevant and personalized content and improve ad relevance, driven more intent to action and focused our content strategy to bring actionable content from a range of sources, including users, creators, publishers and retailers.

As a result, monthly active users have grown more than 30 million over the last 12 months. Our users are also coming back more frequently and are engaging more deeply. Engagement metrics such as sessions, impressions and saves grew substantially faster than our users across all of our regions over the same period.

It's worth noting that we have seen these engagement gains over the past year during a period where we have also grown our ad load on the platform, proving that relevant ads, including lower funnel ad formats can be good content for users in a unique setting like ours with high commercial intent. We continue to see significant benefits and momentum from whole page optimization and ongoing improvements to our ads quality, targeting and relevancy.

We increased our focus on making Pinterest more shoppable by integrating shopping into the core experiences of our platform. We're now seeing strong growth and engagement with shopping related content on our core surfaces. And for the past four quarters, shopping ads revenue has grown multiples of our total revenue growth. I believe we are just scratching the surface when it comes to monetizing lower funnel behavior on Pinterest.

Importantly, we also accelerated innovation on behalf of our advertisers, tripling the number of ad product formats released so far this year versus last and introducing important measurement solutions that prove our value to advertisers. We are seeing revenue from advertisers who take advantage of these measurement tools to grow significantly faster, a positive sign for the future as we increase adoption of these tools among our customer base.

We also opened our platform to third-party demand partners, starting with Amazon Ads platform. To summarize, I'm very proud of what we've accomplished over the last year, and we're just getting started. We look forward to sharing more details and a comprehensive update on our strategy during our Investor Day, taking place in San Francisco on September 19.

With that backdrop, I'll now move on to the progress we've made in the second quarter, advancing our strategic priorities and driving the business forward. First, I'd like to discuss how we are deepening engagement with our existing users and growing new users in Q2, driven by our AI efforts in relevance and personalization as well as a clear focus on satisfaction of user intent.

In Q2, our global MAUs grew 8%, while our mobile app MAUs, which account for over 80% of our revenue and impressions grew 16%. In addition, our U-Can mobile app MAUs grew 9%, accelerating from Q1.

Our users were also visiting us more frequently. The basket of metrics that we use to measure engagement, such as sessions, impressions and saves grew significantly faster than MAUs in Q2 in all regions, consistent with the trends we have observed the last four quarters. To help bring this to life, I'd like to share a few examples of how we are driving engagement on the platform, starting with how we are implementing AI to better serve our users.

Nearly a year ago, we began moving to next-gen AI capabilities, enabling us to use recommender models that were 100 times larger than before. We combined our first-party proprietary data with our AI-based computer vision and search technologies to improve the perceived relevance for recommendations on related pens, driving proceed relevance up by nearly 10 points from a year ago to 94%. This means that when users are searching for something on Pinterest, we are returning results that they find relevant and helpful and we're doing this with a very high degree of accuracy.

More recently, as we have continued to focus on what our users really want from us, we've incorporated into our AI models more signals from our platform. As an example, in the first half of this year, we unlocked further engagement gains by incorporating propensity to share into our AI recommender models.

By recommending content that users are more likely to share, we improved retention of our core users, and we grew revisits from dormant users as they are more likely to engage with recommended content from people they know.

In Q2, another source of engagement gains came from the launch of our new guided browsing experience designed to help episodic users rediscover existing use cases and find their next project or passion on Pinterest. This experience adds a set of horizontal images across the home feeds grid where we can show more content based on the user's intent. The valuable signals we get from this new browsing module helps optimize the user experience for previous or new use cases and deliver more relevant shopping experiences We've seen a meaningful lift in revisitation as a result.

As we continue to focus on improving the inspiration to action journey for users across our platform, we're seeing users explore more interest outside of our core home, food, fashion and beauty verticals, with strong growth in the number of MAUs engaging with emerging verticals such as men's fashion, auto, health and travel in the first half of this year.

Lastly, we continue to see strong growth with Gen Z users on Pinterest who in Q2 were our largest contributor to overall engagement growth and the fastest growing cohort growing double digits and accounting for a larger portion of our overall mix. In particular, we found that our Gen Z users consistently engage more deeply on the platform than our older cohorts as measured by their saves and close up activity on the platform.

Next, I'd like to discuss how we are improving monetization by making Pinterest more valuable for advertisers as a full funnel platform. These ad offerings allow us to reach a broader set of advertisers to diversify and grow our revenue base and also help us deliver improved results to our advertisers based on their performance criteria.

First, let's start with how we are innovating at the lower end of the funnel, where we are helping retailers gain customers, not just transactions. In July, we brought our mobile deep linking product or MDL to general availability, opening up this highly performant offering to more advertisers. MDL is a great fit for retailers who want to drive users to purchase an item on their mobile app and has been a significant driver of our shopping ads revenue growth over the past 9 months.

By taking what we've learned through MDL, we are planning to expand our lower funnel offerings to include more seamless one-click site handoffs in the back half of this year. This gives retailers who don't have a stand-alone app or prefer to do a one-click hand off directly to their site, the flexibility to choose what works best for their objectives.

Travel catalogs are another format we debuted in the quarter, allowing advertisers to reach users who are in the midst of planning, organizing and taking action on their travel dreams. It's built on our product catalog technology and automatically turns each listing into a dynamic product pen with relevant booking information such as hotel name, pricing, images and descriptions.

As an example, TUI, one of the world's leading tourism groups based in Germany added travel catalogs to their full funnel strategy to drive more users to their available hotel listings. In Q2, their campaign yielded strong results with a 6 times higher outbound click-through rate and an 80% lower outbound cost per click than they were seeing in previous conversion campaigns. Furthermore, we made good headway with helping advertisers measure the efficacy of their campaigns on Pinterest.

Since introducing our conversion API in Q4, we've steadily grown the percentage of revenue from advertisers who adopted our conversion API solution. In addition to innovating on the lower funnel, we've also continued to innovate on upper funnel ad formats to build on our strength in that segment of the market and meet advertiser demand.

In June, we made our Premier Spotlight ad format, which is our upper funnel takeover video ad unit on our search surface available for all advertisers. We're seeing early adoption from a large number of brand advertisers such as MAX, formerly known as HBOMax and Comcast XFINITY.

Similar to last quarter, in Q2, we continued to drive a more than 30% increase in our global ad impressions or monetizable supply due to engagement gains and dynamically flexing ad load when a user is expressing intent through whole page optimization. We are also leveraging next-gen AI on our ad products, and we're seeing a profound impact in our ad capabilities.

In Q2, we expanded our use of GPU serving from core engagement AI models to our ad delivery models, which enabled us to use models that are 100 times larger than before in ads as well as organic. The cumulative impact was a 5% reduction in cost per action and over 10% lift in click-through rates.

I'm very proud of our team's execution on the monetization side, but we are still early in our ad platform evolution. I believe third-party partnerships can be additive by expanding the depth of advertisers bringing more relevant ads and improving our auction density.

We announced our first third-party partnership with Amazon Ads in April. This partnership has strong synergies as Pinterest has significant commercial intent and consumer desire to go from inspiration to action while Amazon ads can bring more shoppable content with a great consumer buying experience.

As I said on our last call, a partnership of this scale will be a multi-quarter implementation with the most meaningful revenue impact likely being in early 2024. However, we have recently begun testing live traffic with Amazon ads. We are very pleased with the pace of implementation in Q2 and the early results of our testing in Q3 so far.

Looking at our overall progress on shopping, we're quite excited at the momentum we're generating. We're building a powerful flywheel that is driving increased engagement among our users and delivering greater value to our advertisers. We're utilizing next-gen AI models with our first-party signal to recommend brands and products that are aligned with user preferences, and this is resonating with users.

In Q1, we continued to grow the overall distribution of shopable 10 [ph] or products you can take action on, on the home feed. And in Q2, we've expanded this to all of our core surfaces. Users are actively engaging with this content with click-through rates and saves of shoppable 10 growing over 50%, accelerating from the 35% we saw in Q1.

In Q2, we also launched a new automated shop-to-look module for home decor and fashion pens that utilizes AI to recommend similar shoppable products tied to our merchant catalogs. In our early A/B tests, Shop the - look is unlocking greater value for merchants as this high-intent traffic drove a 9% increase in conversions. And these innovations continue to manifest in robust shopping ads revenue, which continue to grow at a pace meaningfully faster than our overall revenue rate.

Now I'd like to discuss our focus on operational rigor. As we've said in the past, we're instilling a culture where we are more focused on what drives results for users and advertisers and being more rigorous with our expenses, ultimately making us a more durable company. We've been able to accelerate our pace of innovation while also being disciplined on expense management, resulting in positive outcomes for our users, advertisers and shareholders.

In Q1, we took steps to reduce our expenses and drive greater efficiencies in our business, including rightsizing our workforce to align talent to our strategic priorities and restructuring our real estate portfolio. In Q2, we identified further cost efficiencies, leading to operating expenses that were lower than we guided to. Later in the prepared remarks, Julia will share our latest outlook on expenses and margin expansion for the year.

Finally, I'd like to talk about our progress in making Pinterest more positive place on the Internet. Our commitment to fostering a more positive environment resonates with advertisers. We believe our relative brand safety remains a driver for advertisers to invest more and spend on Pinterest. Our efforts and positivity also resonate with users.

For example, we're building inclusive AI technology like skin tone diversification that helps users personalize their experience and see themselves on Pinterest. In Q2, we improved the skin tone diversification of our core feeds, which led to engagement wins. We want others to join us on our mission to make the Internet a more positive place.

In June, we created the inspired Internet pledge in collaboration with the Digital Wellness lab at Boston Children's Hospital. The pledge is the call to action for tech companies and the broader digital ecosystem to unite with the common goal of making the Internet a safer and healthier place for everyone, especially young people.

Before turning the call over, I wanted to say a few words about our new CFO, Julia Donnelly, and why I'm so excited to partner with her going forward. We ran an extensive search for a CFO, and I know Julia is the right leader and partner to help our team grow and scale Pinterest. Prior to joining Julia was VP of Finance and Accounting at Wayfair, where she was responsible for the entire global finance and accounting organizations as well as Investor Relations and Corporate Development. She is a strategic, disciplined and highly regarded executive with an impressive background leading all aspects of finance within an innovative, high-growth public company. Her leadership and deep experience across e-commerce, media and technology will be invaluable as we position ourselves for the next chapter of entrust.

I'll now turn the call over to Julie to provide an update on our Q2 financial performance.

Julia Donnelly

Thank you, Bill, and good afternoon, everyone. I'm thrilled to join the Pinterest team. As you can imagine, it's been a busy and exciting few weeks since I joined in late June. First, I'd like to discuss a few of the reasons why I came to Pinterest and why I'm even more excited about the opportunity ahead.

Before joining, I followed Pinterest extensively as a beloved brand with an inspiring and positive consumer app and a strong financial model. From my previous role, I've also watched Pinterest become an incredibly performant and innovative advertising platform.

Coming from what I would consider a best-in-class e-commerce advertiser, I've had first-hand experience and how Pinterest is uniquely able to help marketers find users who are in the midst of an inspiration journey where they're looking to buy but not quite sure what they want and then ultimately take action to bring that to life.

Over the last few weeks, sitting inside the company, I can attest to the excitement and energy from our teams on everything we are building for our users and advertisers. I'm honored to take on the CFO role and join the management team at Pinterest to continue building on this momentum. Leveraging my background in e-commerce, I look forward to helping us navigate through this exciting next chapter of growth, capturing the significant revenue potential ahead through strong operating execution and sound investments.

Looking ahead, I'm excited to develop a productive dialogue with the shareholder and analyst community as well as sharing more about progress to date and future potential ahead at our Investor Day in September.

Now turning to our results in the second quarter. In my remarks today, I'll discuss our Q2 financial performance and provide our preliminary outlook for Q3. All financial metrics, except for revenue will be discussed in non-GAAP terms unless otherwise specified, and all comparisons will be discussed on a year-over-year basis unless otherwise noted.

As Bill mentioned, our investments in AI and new user experiences drove strong engagement on our platform in Q2, resulting in 465 million global monthly active users, an increase of over 30 million users over the last 12 months. We grew monthly active users 8% on a year-over-year basis globally, the highest growth we've seen during the last 2 years.

In the U.S. and Canada, monthly active users were 95 million, up 3%. In Europe, monthly active users were 124 million, up 6%. And in our Rest of World markets, monthly active users were 246 million, up 10%. Underlying our monthly active user metric, we had even faster growth and engagement as measured by sessions, impressions and SAFE [ph] which is a testament to the work we're doing around deepening engagement with our users. And I'm excited about the progress we're making with Gen Z users, which was our fastest-growing segment and an important and influential demographic for marketers in many of our core verticals.

Turning to our financial performance. Global revenue for the second quarter came in at $708 million, growing 6% or 7% on a constant currency basis. U.S. and Canada revenue was $565 million, an increase of 4%. Europe revenue was $114 million, growing 12% on both a reported and constant currency basis.

Revenue from Rest of World was $29 million, growing 32% or 35% on a constant currency basis. We're pleased to see global year-over-year revenue growth continued to strengthen this quarter versus Q1 '23 and Q4 '22. A testament that while the ad market remains somewhat choppy, advertisers continue to turn to turn to Pinterest because of our highly attractive user demographics, relative brand-safe environment and the strong return on ads spend we can deliver.

During the quarter, at the top of the funnel, we saw strength and awareness objectives as many brand advertisers, particularly CPG advertisers, were able to meet their goals given our engagement gains and attractive pricing. Contributing to this growth were video ads and our new premier Spotlight takeover ad unit.

At the bottom of the funnel this quarter, strength came from larger, more sophisticated advertisers who leaned into our shopping ads format driven by adoption of mobile deep-linking and our conversion API. However, we also experienced headwinds from those who have yet to adopt our measurement tools and pockets of distressed retail advertisers.

Finally, across all parts of the funnel, we continue to gain good traction with advertisers in emerging verticals such as travel, autos and financial services, helping us further diversify our advertiser base outside of our traditional areas of strength.

In terms of overall auction dynamics in the second quarter, ad supply grew more than 30%, driven by engagement gains and opening up supply to whole page optimization, and the price of ads declined 20% due primarily to a softer pricing environment across all geographies.

Moving down the P&L. Cost of revenue was $164 million, up 1% year-over-year and down 2% quarter-over-quarter versus Q1. This is the third quarter in a row of sequential declines in our cost of revenue line item as our teams implemented a number of efficiency projects across storage and compute infrastructure. It has been exciting to see how we can realize these efficiencies in our cost of revenue even while we drive significant engagement gains and ad stack innovation and leverage AI throughout our business.

Overall, our non-GAAP operating expense was $440 million, up 5% year-over-year and up 7% sequentially versus Q1. This was below our guidance as we were able to drive more cost efficiencies, particularly in our R&D and G&A functions while also shifting some of our marketing spend from Q1 into Q2.

Adjusted EBITDA had a very strong quarter, coming in at $107 million or a 15% margin, up 130 basis points year-over-year, driven by increased revenue, efficiency in our cost of revenue and operating expense discipline. This quarter, we successfully completed our $500 million stock repurchase program, well ahead of the 12-month plan to complete the program as we saw openings to be opportunistic with our purchases during the quarter. We continue to be mindful of managing dilution. Capital allocation will be a topic of our - at our upcoming Investor Day, and we look forward to sharing more details at that time.

Finally, we ended the quarter with approximately $2.3 billion in cash, cash equivalents and marketable securities.

Now I'd like to provide preliminary financial guidance for the third quarter. While the overall digital ads market appears to be stabilizing, the recovery remains uneven with pockets of strength and weakness. We expect Q3 revenue to grow in the high single-digits range year-over-year. Based on what we are seeing right now, we expect to land towards the middle of that range, which would represent a moderate acceleration from the growth rate we saw in Q2. This includes a small impact from foreign exchange, which is reflected in our guide.

Turning to expenses. We expect Q3 non-GAAP operating expenses to grow in the low single digits range year-over-year. Please note that our operating expense guidance does not include cost of revenue. As we have noted previously, we are committed to expanding adjusted EBITDA margins in full year 2023 on the order of a couple of hundred basis points year-over-year in any state of the world.

Based on the revenue acceleration and the expense outlook provided above, we believe the adjusted EBITDA we expect to deliver in Q3 puts us on the path to double that prior commitment, resulting in roughly 400 basis points of year-over-year adjusted EBITDA margin expansion for full year 2023. We feel good about the progress we've made on users, engagement and monetization while also maintaining cost discipline to ensure we are investing in areas with clear line of sight to strong returns to capture the sizable revenue potential we see ahead.

With that, I'll now turn the call over to Bill for a few final words.

Bill Ready

Thanks, Julia. I'm proud of our team's execution to deliver strong results in Q2. And I'm confident that we're making the appropriate investments to provide great experiences for our users and more value for advertisers. I want to thank our teams at Pinterest, our advertising partners and all the people that come to Pinterest to find inspiration.

And with that, we can open the call up for questions.

Question-and-Answer Session

Operator

We will now open the line for questions. The first question comes from the line of Eric Sheridan with Goldman Sachs. Your line is now open.

Eric Sheridan

Thank you so much for taking the question. Maybe two, if I could. Bill, first for you. You talked a lot about product and platform iteration and there's also a lot of talk about the current advertising environment. Can you talk a little bit about when you think about the elements of revenue reacceleration that are building in your business, how much are elements that are within your control based on things you're building and sort of executing on inside the business versus elements of volatility down to the broader ad environment or macro environment that are somewhat out of your control.

And then the second one, Julia first, congrats on the new role. I look forward to working with you going forward. I wanted to come back to your comments during your prepared remarks and just delve a little bit deeper in what you see as some of the big opportunities or agenda items that you want to tackle in this role as the CFO? Thanks so much.

Bill Ready

Thanks for the question, Eric. So to your first question, on what we can control and what's sort of in the broader environment. The broader environment is seeing some stabilization and some signs of recovery, but there's still a lot of puts and takes and a lot of choppiness out there. Direct example of that is if you look at pricing impact, which would be sort of a direct evidence of sort of softness in the ad market, that's down 20% year-on-year for us. That's pretty consistent with what you see across other large platforms, some of which have had year-on-year declines last year and year-on-year declines again this year, they were double digit plus or 15 percentage point plus.

So you have this pretty strong pricing headwind, which is just the element of like softness in the broader market. And so while there is some signs of stabilization in pockets of recovery, there's still choppiness in that as evidenced by both us and other major platforms with that pretty material headwind on pricing.

But if you look at - to your question of what we can control there, we feel fantastic about the progress that we're making. And the things in our control, most noteworthy, the return to user growth, we just put up our largest user growth quarter in 2-plus years deepening engagement on the platform. We've talked about how, given the high commerciality of what users do on our platform, that we can take ad load much higher than what it has been that ads in a commercial context can be really relevant content. And so we can take ad load higher even while driving deeper engagement, and you see that evidence in our results.

That's a really unique thing in the ad world to be able to do that. And particularly in social media, where ads tend to sort of take the user away from what they're trying to do because they're more of an entertainment mode. Here, because we have the user in a lean forward intent mode, we are demonstrating through our whole page optimization work and our dynamic ad load that we've been able to take ad load up 30% - or drive 30%-plus lift in impressions, part of that driven by growth in user engagement, but a big part of that driven by increasing ad load while engagement is up meaning we're proving out those ads as relevant content.

So I think that, coupled with the advertiser side of this, where as they're getting more adoption of our measurement tools, they continuously as they adopt our latest measurement tools, spend more and tend to be much faster growers than those on the rest of our platform. I've shared this in some prior remarks later in the quarter this past quarter that those that have adopted our measurement tools tend to grow on the order of 30% year-on-year. Those that have not yet adopted our mid-single-digit decline. So as we're driving that adoption curve, that headwind will increasingly flip around to tailwind for us.

And so those are things that are in our control that again, if you step back from it, give us lots of confidence in where we're going. And I think it's why if you step all the way back and you look at us versus the broader ad market, we're one of the only major ad platforms that has grown consistently every quarter through the downturn in the ad market. and why we're seeing acceleration quarter-on-quarter.

And while that acceleration is moderate, when you look at the 20 point-plus headwind on pricing and the fact that we're one of the only consistent growers through the downturn, I think that speaks to the really positive progress on the platform. And then final point, I'd just say a year ago, one of the big questions was would shopping work on the platform. We had user research to send more than half the year here to shop, the wood shopping work on the platform. And there, I think our results just could not be more encouraging.

We talked about 50%-plus year-on-year improvement in engagement with shoppable content as we brought shoppable content on to our main surfaces. And that's accelerating quarter-on-quarter from 35% last quarter. So that as an indicator of the commerciality of our platform and what we can do for advertisers, I think, again, just gives us tremendous optimism. And while we can't predict when the sort of pricing headwinds in the broader ad market alleviate or turn back into a tailwind, we've been able to deliver really solid growth even with a 20-point plus headwind. And I'll give it to Julia for your second question.

Julia Donnelly

Thanks, Eric, for the question. I appreciate it. So first, let me start with just a couple of observations. It's been obviously an exciting few weeks here for me at Pinterest. And from my prior seat, I heard a lot about Pinterest and the exciting pace of innovation that was taking place.

Now that I'm on the inside, it's great to see that many other advertisers are really recognizing the same. When I first got here, the teams had just gotten back from can. They were talking about the buzz and momentum from agencies and advertisers about all the ad products we're building. As we mentioned earlier, we've actually tripled the number of ad formats released so far this year versus last year, and we're really accelerating the pace of ad stack innovation meaningfully as we described earlier. So that's starting to really get recognized by the broader advertiser community, which is great to see.

The second observation I'd have is just how impressed I've been early in here just the level of engineering and product talent that we have inside the company here, particularly in our AI functions. We have really world-class AI teams that have - Pinterest has been a talent magnet on that front. So I'm impressed to see and learn how embedded AI actually is in virtually every aspect of our platform today.

And as you think about sort of agenda items and where we go from here, obviously, I'm very excited about the direction that this team already has taken a company and that's a big part of why I chose to join and help build on that momentum. I'd say we've made some real progress over the last year. But my - where I look forward to partnering with the team is continuing to make sure we're driving those investments in areas with really clear line of sight to returns where we're investing in that growth to capture the revenue opportunity but also doing so profitably and looking to drive that operating leverage over time.

Operator

Thank you. The next question comes from the line of Ross Sandler with Barclays. Your line is now open.

Ross Sandler

Hey, guys. So we're starting to see these really nice SKU-level ads from Amazon showing up in the app. And you just mentioned that you guys have been able to increase that load pretty nicely even before turning these on. So I guess, can you just talk about like how you think about ad load, the ceiling to ad load if there is one? And then how these new Amazon - compare from a pricing perspective to the directly sold interest ads?

And then the second question is just on the new categories, the travel, financial services, auto, just how you guys are viewing? I know it's early, but how are you viewing those non-endemic categories in terms of potential opportunity in the next couple of years? Thanks a lot.

Bill Ready

Thanks, Ross. So on your point on the Amazon Ads partnership, as I've said before, we really believe that the commerciality on our platform can allow us to take ad load much higher. We don't break out ad loads specifically, but we've talked about the 30%-plus growth in monetizable supply, which is meaningfully faster than our engagement growth overall. That's really been driven by bringing more relevant shoppable content onto the platform for users.

So Amazon as a partner, we're quite encouraged that they can bring more of that shoppable content plus a great buying experience for users. And so while it's a multi-quarter implementation and again, a reminder that the biggest element of that revenue we expect to start hitting in early '24, we're really pleased with the pace of implementation. And in those early results, if I explained a bit on those early results that we're so happy about one of the biggest things we're happy about is that we're seeing improvement and relevancy even beyond what we're already very optimistic expectations on our side.

And so as we see that it's contributing to more relevant shopping content that further bolsters our view that we can take ad load much higher. And if you compared to other analogs that I've seen in past life and that you all could see as you sort of count ad load or add slots on other sort of shopping experiences. -- ad load provided we have the right relevancy, Adlocan [ph] be a multiple of what it is today for our users particularly when users are in commercial context and more than half the users are here to shop. So I think that says that there is still a lot more runway ahead, quite a high ceiling for where we can go with ad load.

And the fact that we've already been growing at 30% plus year-on-year in terms of overall monetizable supply, while also increasing engagement is proving out that theory of the case that we can grow ad load while also making it relevant engagement driving content for our users. So we feel really good about all that.

And then to your second question on new categories and how we're seeing those sort of previously non- categories, across Pinterestinc, we talked about the uniqueness of Pinterestinc being that it's a great full funnel platform where people would go from inspiration to action. But we knew that actionability was historically a weaker part of the platform as we've been bringing in more satisfaction of intent, more ability to go from intent to action, we're seeing that, that is resonating across categories, across verticals. And so a few that we mentioned, like travel, auto, men's fashion, we're seeing that as we're doing more to create satisfaction intent, that is applying well across other categories. Travel is probably the most noteworthy I mentioned in my prepared remarks, travel is quite a large ad market overall. People have always dreamed about travel on Pinterestinc.

But again, like shopping, the actionability was lower as we've taken our work in shopping, applied a lot of that base technology to travel and opened up a travel catalog as we now have made more of those travel ads and clickable and engageable by the user with things like pricing information availability we're seeing that, that intent to action effort is driving really good results with users and with advertisers.

So I think that's quite encouraging that while we still have tremendous runway to go in shopping, there are other meaningful categories and verticals on Pinterest. And so we're quite excited about that. Finally, I would just say on the organic side of that, I mentioned some of what we're doing around helping to nudge users into other use cases. I think this is one of the things we've seen historically is that Pinterest could at times be more episodic for some users, and we've been leveraging some of our AI capabilities and some of our product experiences to now nudge users towards other relevant categories for them. And we're finding that is working really well. So we think there's a lot more runway to go there.

Ross Sandler

Great. Next question.

Operator

The next question comes from the line of Lloyd Walmsley with UBS. Your line is now open,

Lloyd Walmsley

Okay, thanks. A couple for me on the partner monetization side, if I can. First, it sounds like Amazon partnership is progressing faster than you guys expected. What are the big outstanding blocks you need to get through to scale that? And as you think about potentially adding other partners, like are those things that would need to get redone from scratch? Or is this more foundational platform efforts?

And then second one, the partnership you guys announced with Alf [ph] how meaningful could that contribution ultimately grow to and in markets where you have their sales force helping you and other third parties, how will you kind of toggle between those sorts of go-to-market and a less? Thanks.

Bill Ready

Thanks for the question, Lloyd. So on the partnership with Amazon and sort of your broader comments on bringing on third-party demand, as we said, yes, we are very pleased with the pace of implementation and even more importantly, pleased with the significant improvement in relevancy and what that says about the long-term potential of that partnership.

And so a couple of things I would note. You asked about the pace there, yes, we are pacing ahead of expectations. At the same time, as we get relevancy really right that places a much higher ceiling on the long-term potential of the partnership. So in the near term, we're making sure that even though we're really pleased with the pace and the progress and the results, then we want to make sure we do that in a methodical way, getting the relevancy really right for the user because as we do that, that gives us a much higher long-term ceiling. So we're not going to sort of sacrifice the long term to sort of go faster than is necessary in the short term, but we are seeing really positive progress there. And I'd say our long-term view of that partnership just continues to expand, even though we want to be careful not to get ahead of ourselves in terms of revenue contribution for this year. We do think that when you step back and say, our big opportunity here is about making our platform shoppable, in content to users. What we're seeing there is extraordinarily encouraging. I think the ceiling only continues to lift as you look at the long term there. And the pacing again is better than where we had expected.

You had asked about other partners. We've mentioned this previously, we do expect to have multiple partners to help us bring in engaging relevant content, ad demand for our users. We're focused on making sure we get this first one really right. But we are building in a way that will help to sort of agree the skids for others. And so we do think as we land this one really well, and that's tracking well and we start to look towards other potential partners that we're building this in a way that is repeatable and will help us continue to accelerate as we think about where we go forward.

But again, we're going to make sure we get this one really right, particularly getting that user relevancy right because we get that right, then the ability to go take that ad load up much higher while still deepening engagement with users is really the longer-term potential here. And again, all signals there are encouraging even beyond our prior expectations.

You asked about what we're doing with LF [ph] and how meaningful that could be. So for that, as we think about international, we do think that our international monetization is a big opportunity for us. This past year, we had to really focus on making sure that we made some meaningful updates to our user experience and our ad capabilities in our largest markets, which are highly competitive, and so we feel good about the progress we've made on that.

We did start to make some focused investment in international, particularly with agencies and partners that can help us drive more distribution and meeting our advertisers where they are. And so as we have leaned into agency deals and partners that can help us distribute internationally, we have seen those have performed quite well without speaking to any one specific partner, we've seen that those have been growing at a much faster rate of revenue than what we would see otherwise

So we think there's a lot more yield in that. And as we look into '24, we think international that has probably been under monetized across the board, we have been even more under monetized internationally. And so I think we have some really encouraging proof points to say there's a lot more we can lend into international monetization as we go into '24, particularly with partners like olefin others that can help us accelerate that distribution. Operator, next question.

Operator

The next question comes from the line of Mark Mahaney with Evercore. Your line is now open.

Mark Mahaney

Okay. Thanks. I want to ask about the two things: the LLM models, the impact you've seen in terms of cost per action, reduced cost per action and lift in CTRs, I would assume these are still kind of early stage results for you. Could you talk maybe about what kind of long-term impact you could see? Or is that kind of in the unknown knowns. How much better do you think the application of large language models to Pinterest much more of an improvement could you see on both the user and the advertiser side?

And then one specific question about this data point you provided on a month ago about how 10% of your advertisers were -- had adopted your measurement tools, and they're growing 30% year-over-year in terms of their ad spend with you. Where are those two numbers now? And like what's a reasonable expectation for when 10% gets to 50%? Thank you very much.

Bill Ready

Thanks, Mark. Appreciate the questions. So on LLM models, two things are simultaneously true. O ne is we are seeing tremendous benefit, and we are pretty far along with what we're doing there. The second thing is true, though, is that as you look at the longer, I think there is a lot more runway to go in terms of what we can do with that core technology, particularly when you consider the fact that, that technology is going to continue advancing.

So to be more specific, on larger models and next-gen AI capabilities moving from CPUs and GPUs, things like that. I mentioned in my prepared remarks, we really started into that nearly a year ago. These things were available prior to sort of a lot of the sort of big uptick in sort of the news cycle around earlier this year. These things were available, and I worked close with them in past lives, knew the potential there, and it's something that we really started to lean into nearly a year ago. We started first on the organic side.

And so as you look over the past several quarters, the fact that we've been able to take our recommender models up to be 100 times larger than what they were. I mentioned on the call that 10 percentage point improvement in perceived relevancy by the user. In the world of search, advertising, discovery, 10 percentage point improvements in relevancy are like completely unheard of, like the magnitude of that like cannot be overstated. So I think that says, okay, we're seeing really tremendous impact from that, and we're well on our way on that on the organic side.

On the ad side, we just started bringing that over to the ad side now with models 100 times larger. So we got right on the organic side first. Now we're bringing it to the ad side. And yes, same thing on the ad side, when you look at those things like a 5% reduction in CPA and a 10% lift in click-through rates, again, in the world of advertising, those are unheard of kind of numbers.

And so we're early on there, but the impacts were profound. So we're quite excited about how much more runway is left with those things. But we're deep enough in it to know that we feel like we are absolutely drafting on that acceleration in innovation that's happening in AI generally. It's been a core strength of Pinterest for a long time, computer vision as one of the places that has been sort of at the forefront of sort of advancements in AI and computer vision has been a strength for Pinterest, and we've got a really talented team there. So we feel really well positioned on continuing leverage from AI or further along on organic, but with a lot of runway to go on the ad side, we're earlier there, but already seeing profound impacts. So again, tons of runway on both.

I think on your second question around the 10% of advertisers that had adopted measurement tools growing 30% on the range of 30%. We gave that update pretty deep into the quarter last quarter. So we're not updating those specific numbers at this moment. We're going to have our Investor Day coming up. We'll talk more about some of the progress there.

The one thing I will say is that we've seen really nice continued adoption. The biggest thing I've been watching for is, as we continue along that adoption curve, do we continue to see the linkage between advertisers adopting the measurement tools and then seeing that they were getting much better performance than what they had realized from Pinterest and increasing spend. And that linkage, not only has that remained present as we've continued along that adoption curve, that linkage is becoming stronger.

And maybe to give a little bit of a qualitative example of that when we were at can [ph] not that long ago and great gathering of many of the world's leading advertisers, we had tremendous buzz there on a number of fronts, just around the overall acceleration in our platform, our users and the acceleration of our ad platform.

But this point of advertisers adopting and seeing that they were getting much more performance than what they had realized and there were much more performance oriented solutions available for them than what had been available for things like mobile deep link and things like that, that are driving great lower funnel performance. There was really good buzz around that. And there's nothing better than having a large sophisticated advertiser out there saying, "Oh, yes, we adopted that from Pinterest and saw really great results.

And I think I mentioned that in my prior commentary that as we look at those that have adopted so far or the early adopters actually tended to be the larger, more sophisticated advertisers, the hardest ones to please.

So I said before, if you can make them happy, you can make anybody happy. And that's what we've been seeing is that as successive cohorts are adopting, that linkage is not only still present, but getting stronger in terms of them seeing more performance than they realize and therefore, increasing their spend with us.

Operator

Thank you. The next question comes from the line of Rich Greenfield with LightShed Partners. Your line is now open.

Rich Greenfield

Bill, you were talking about sort of the AI and targeting. And I guess there's just this growing fear among investors that everyone not named Meda [ph] Google, Apple will be sort of a little struggle to compete in AI as we look out over the next 2 years? And given the importance of AI bot as you were talking about from content recommendations and ad targeting, and I know sort of the history you spoke to at Pinterest in terms of computer vision. But would love to get a sense of how you think about how you compete with the - what is, I guess, the juggernaut in tech, especially given your background at Google. And then I just got a quick follow-up on demographics?

Bill Ready

All right. Maybe we get the follow-up quickly, just so we know what it is, and we don't lose it from you.

Rich Greenfield

The follow-up is just you talked about the demographic group, and you talked about Gen Z growing so quickly. As we look at all of the different demos that use Pinterest, especially as you think about shopping, the actual -- taking action on content and buying content, do you see any major deviations or changes in terms of what behavior looks at by demographic as you look by cohort? Anything to kind of call out in terms of what you're seeing on shopping activity specifically?

Bill Ready

Okay. Great. Thank you, Rich. So on your first question on AI, I think there's a question of do the benefits only accrue to the largest. I think we're already providing a pretty strong counterpoint to that. And actually, if you step back, I think if you look at like cloud computing, while yes, cloud computing had a lot of benefits to the largest providers of cloud computing, it also created a lot of ability for a broad swath of people to engage.

And I think you're seeing that with AI that yes, you have some of the largest players that are providing a lot of great capabilities there, but they're also externalizing those through cloud compute. It's also the case that the open source community is advancing really rapidly, and you're seeing some of these open source models are actually comparing quite favorably to some of the things that you would have seen come from the very largest players.

And if I step back for a moment, I would actually take the view that I think in this sort of first round of next-gen AI, I think you've seen a lot of value accrue to the model creators. As you play that forward, I actually think there's going to be a lot of value that starts to accrue to the places that have really unique signal and really unique user interactions.

The AI is only as good as the signal of home, which is acting. And when you look at it through that lens, we have a really unique signal at Pinterest. We've talked about this that unlike the rest of social media, we have very high commercial intent. But if you go further in that, it's not just we have commercial intent, it's not just that we know what people might click on or buy. These commercial journeys are long-lived, and users on Pinterest interestingly, are curating for the future. So we don't just know what people have done in the past. We know what they're planning to do in the future

. When you think of the power of that signal and what that can do in terms of the power of recommendations that we can make to users, the advertising capabilities that we can open up with that I think we just have a lot of uniqueness in our signal from the fact that it's not just the people who are shopping here, they are curating here. They are not just saying, okay, here's the pair of shoes I'm interested in the hand value that I'm interested in or the dress that I'm interested in, they're saying, here's how I think about putting a great outfit together.

And this handbag and these shoes go with distress, and that's being repeated across hundreds of millions of users on our platform. That is a completely unique experience that doesn't exist anywhere else, certainly not in the Western world. And that really unique signal as we're pairing that with those models, I think that's letting us do really unique things with the way that we can make recommendations.

So when I talk about that relevancy, a really interesting thing to think about for us it's not even just that a search relevancy, you're saying, "Okay, you search for the sign we give you the right thing. We're doing that on things like related items. And so to get to things like 94% plus relevancy, when we're doing broader sort of discovery for the user, that is exceptionally difficult, but is an example of us taking the models that are becoming more generally available for all and then applying them to really unique signals and really unique user interactions on our platform.

So I feel that we are quite well positioned in the longer term sort of move towards next-gen AI and the dynamics there, I think, are quite favorable for us. And again, I think no better way than to demonstrate that in your results, and we've been doing that.

I think your second question on demographic groups. We've noted this multiple times that Gen Z is our fastest-growing demographic and our largest contributor to overall engagement, which means it's not just a high growth rate on a small denominator, it's our largest contributor to overall engagement growth. That's a really interesting thing I've worked on a lot of different consumer products and a lot of different platforms. Generally, as you get a more mature platform or a more mature product, each successive cohort is a little less engaged, a little weaker.

To be able to say that we have Gen Z a more recent cohort that's actually engaging even more deeply than our prior cohorts, I think is quite exceptional. And I think it gets to the fact that what we're solving with this ability to go from dreaming to doing and to use pents as a way to play in your future, whether that's in travel or in shopping I think that's really resonating in a lot of our next-gen experiences that we're bringing are resonating there as well, things like what we've done with shuffles and the ability to go even more uniquely curate these things, that's working quite well.

Shopping , you asked about that specifically. I think we're finding that to be broadly resonant. And so not only is the shopping behavior broadly resonant across demographics, I mentioned that, that shopping behavior or sort of if you step back and say, doing more inspiration to action, we're seeing that work across other verticals and categories. That's quite encouraging for us as well like the travel sampling we've spoken about. Operator we'll take our last question.

Operator

Absolutely. The last question comes from the line of Doug Anmuth with JPMorgan. Your line is now open.

Doug Anmuth

Thanks for taking the questions. Bill, I know you touched on Tapi [ph] I guess I just want to understand what's keeping advertisers from adopting here? And what are you doing proactively to move them along the adoption curve?

And then secondly, any comments on the MAU decline in Europe in 2Q. I don't know if that was part of the birthday collection dynamic that you've talked about. And just curious if you have any thoughts on the overall 3Q MAU outlook? Thanks.

Bill Ready

Thanks, Doug. I'll take your first one and then give Julie the second one. So our conversion API is important to hit. Our conversion API, we've only recently put that out there. It's been on the market less than a year. So we're early on with that. So to say what's keeping folks from adopting it, we're really pleased with the progress that we've been making on that. It's a recent product for us, and we've seen really significant uptake of that product. So we feel really good about it overall.

But of course, we want to be faster, given the magnitude of its impact is so consequential for us and the advertiser. I think the spirit of your question of like, hey, how do we move along the adoption curve faster. We very much are mindful of that. So a few things that we're thinking about.

The first is with that being a recently introduced product, we've really been getting our own go-to-market really tight in terms of how our sales team is engaging. The more we have proof points, I talked about having large sophisticated advertisers that are speaking about how impactful it's been for them. That - those reference points are really compelling that's really helping us drive better and better go-to-market with our first-party efforts there.

But the other thing that we're doing is working to meet the advertisers where they are because there's a broad-based adoption curve on these things, it's not just us with the larger platforms are putting these things out there as well. And so advertisers are moving through an adoption curve on this. So we're making sure that we're working with third-party partners. We've mentioned some of them like LiveRamp and Tealium or Google Tag Manager and then certainly, our recent partnership with Amazon's ad platform, these are all examples of integrating to the platforms where those advertisers are already integrated.

So that sort of multipronged approach of saying we're going to get better and better with our own go-to-market, which is working really well and advancing quite nicely as well as meeting the advertiser where they are. Those things are sort of the two-fronted approach of how we're accelerating that. And again, the most encouraging thing is that consistently as those advertisers adopt these measurement solutions, they're seeing better performance than they expected from us and increasing their spend with us.

So we think that is right now, the balance of those advertisers given us a new product being tilted towards not yet having that tool while that creates some headwind for us. We talked about that cohort being a mid-single-digit decliner. But the ones that have adopted those tend to be growers on the order of 30%, that headwind flips around a tailwind as we get more of that balance of adoption. And we are seeing strong pull-through on that.

Julia Donnelly

And then, Doug, just on your second question about lumpily active user kind of trends, specifically in Europe and globally and now look for Q3. I'd say, as Bill noted, we feel really great about the monthly active user growth that we've seen globally. So we're at a 2-year high, both globally and in the U.S. and in rest of world and near-tear high in Europe as well, if you look at year-over-year.

Comps on MAUs. So I think we're feeling really good there. There is some seasonality in the business on monthly active users as we called out on our prior earnings call. This is Q2 is a 30-day look back and during the month of June. And so you do typically see a little bit more seasonal weakness in Q2 and then that reverts back in Q3. So it's important to really focus on year-over-year metrics looking here. And on that basis, we're continuing to feel quite strong. Birthday collection was something we noted on the prior call that ended up being a nonfactor and not material.

So overall, feeling very good about all the trends that we're seeing on MAUs, the return on our marketing spend and all the user and growth engagements that we're investing in. And I would expect that we don't guide on MAUs for Q3 specifically. But in general, we feel really good about being able to maintain those year-over-year increases going forward.

Neil Doshi

All right. Thanks again to all of you for joining the call and for your questions. We look forward to keeping the dialogue going, and hope you all enjoy the rest of your day. Thank you.

Operator

That concludes today's call. Thank you for your participation. You may now disconnect your lines. +