Good morning, happy Wednesday morning of the 2024 Morgan Stanley TMT Conference. We're thrilled today to have Bill Ready, the CEO of Pinterest, with us.
Thanks for having me.
Welcome. Welcome back. Good to see you as always. The real way to wake up is from some disclosures, so we'll get started there.
All right.
Important disclosures, including the Morgan Stanley research disclosures, are available at the Morgan Stanley website, www.morganstanley.com/researchdisclosures. They're also available at the desk. Some of the statements that Pinterest will make today may be considered forward-looking. These statements involve a number of risks and uncertainties that could cause actual results to differ materially. Any forward-looking statements that Pinterest makes are based on assumptions as of today, and Pinterest undertakes no obligations to update them. Please refer to Pinterest Form 10-K for discussions of the risk factors. They may impact actual results. So you're coming upon 2 years-ish in the job. There's been a lot going on at the company in the last two years and even the last six, seven months with Analyst Day and kind of developed since then. So maybe let's talk about that.
So if we sort of say, take me from when you started to the Analyst Day, then the Analyst Day to now, and how your priorities and kind of what you're focused on most has changed to sort of drive engagement and monetization and the platform's potential.
Yeah, great. Well, you know, so middle of my seventh quarter right now. It'll be two years, end of June. A couple of years ago, as you looked at Pinterest, users were declining. Revenue had gone to sort of mid-single digits-ish, and margins had compressed significantly. There were people asking about, like, hey, does Pinterest kind of survive in a TikTok-driven world, right? Those were the questions. Some people would say, like, so Bill, what are you doing? While some folks saw that, what I saw was that Pinterest had a very unique use case, a tremendous amount of commercial intent that had been really untapped and could be a very different kind of ad platform than what it had been previously.
If you look at sort of where we were versus where we are, you've seen us accelerating on revenue through the ad downturn, where only the public platforms that grew consistently through the downturn. You saw us accelerate in the back half of last year, back to double-digit growth on revenue, as you can see implied from our Q1 guide. We see further acceleration as we come into this year on the revenue side. But that has really stemmed from really getting it right with our users. You look at the user acceleration, so from declining users to accelerating user growth with our basket of engagement metrics growing faster than users.
So even as we've accelerated the growth of users, we've deepened engagement per user, which is oftentimes really hard to do as you expand users, especially later into a product, because normally you get your most engaged users first, and each successive cohort's a little less engaged. Well, we have deepened engagement per user. In fact, our most recent user cohort, as we shared at our Investor Day, roughly twice as engaged as prior cohorts had been. So we're finding our best product-market fit in years. And at the core of that has been great execution on AI, using that to drive big improvements in relevancy and big improvements in actionability on the platform, which had been sort of the Achilles heel of the platform previously.
So that's what's driven up the user growth, user engagement, and married that well with the revenue side of the business, which has shifted from primarily an upper-funnel brand ads company two years ago to now roughly two-thirds of our revenue is lower funnel. That's a massive shift in the business. That's a, from my background, a much larger, more attractive part of the ad business to be in. So we've made a big shift in that. So all that is really great progress. Then just the executional discipline of the company, the operational rigor, we went from contracting margins to last year we promised 200 basis points of margin improvement. We ended up delivering 660 basis points of margin improvement. So like to underpromise and overdeliver on those things. So I think we're getting a good, solid track record on those.
And so you then said, well, OK, that's sort of two years ago to today, Investor Day to today. What I would say is that what we laid out at Investor Day, and we talked about our three to five year view of what revenue and margins could be. We talked about the sort of multiple ways to win and the multiple levers there. What I would say from Investor Day to today, and again, you see it implied in our guidance, we see really great progress toward that plan. That plan is very much playing out as we intended across each of those key levers that we talked about. We see real progress on each of those. So we feel really great about where the business is headed. And obviously, there's tons more execution to go.
There's, as much as it's been transformative the last couple of years, it's a great start, not a finish. And there's tons more untapped potential, right? When I talked about the untapped commercial intent in the platform, as much progress as we have made, we are still really just getting going with that, right? We've talked about how shopping is a first vertical that we're focused on. We've talked about other emerging verticals like travels and autos and financial services and all these things. So in due time, there's a lot more we can do to just drive better and better expansion of adjacent use cases with users. So we're really focused on nailing shopping first. But that depth of engagement per user, you see that really, really working well.
So again, if I had to pick one highlight out of all that stuff, the fundamental question was, can you really make the product relevant for users? And can you stand apart in a world of everybody running a short-form video, TikTok, and Instagram, and YouTube Shorts, and all these things? And what I saw was not that we needed to go fight on that battlefield, that we just had a totally different business to build. And that's why you see us re-accelerating with users. And one last thing on just how good the product-market fit is. I've seen a few folks comment on this. It is exceedingly rare to see an app age down. Normally, it's, OK, your user base ages. And then, oh, somebody else comes in and captures the next generation, and you sort of age out or age up. We've aged down, right?
Gen Z is now more than 40% of the users on our platform, our largest fast-growing demographic. So just again, if there's one highlight, it's around just what we've done to make this a great product for our users that is distinct and separate from what's happening elsewhere in the ecosystem. And there's a ton more of that to go.
Got it. OK. There's a lot to unpack there. So maybe let me go back to sort of the Analyst Day, where you actually laid out, I thought, a pretty broad menu of areas you were focused on, execution for the user side, both relevancy, ways to drive engagement, and then all the ad tools from the API, Direct Links, deep links, et cetera.
Maybe now, here we are, six or seven months after the Analyst Day, talk to us about areas that have gone better than expected on that sort of list of execution points and areas that have perhaps proven to be more challenging from a micro-level perspective on those.
Well, what I would say, again, each of those levers of growth that we talked about, we see really strong progress against each of those. So we talked about users, a bunch. On the ad side, we're driving through an adoption curve to lower funnel. And while we've had great progress on that, I talked about now roughly 2/3 of our revenue comes from lower funnel versus the business being almost entirely brand ads driven a couple of years ago. It was still the case that we're early in that adoption curve. I shared this stat on the last earnings call. At the start of last year, because our lower funnel tools were brand new, at the start of last year we had 2% of our customers and our revenue coming from that had at least three of our lower funnel tools. By Investor Day, that was 13%.
By this last earnings call, we reported out that it was 23%. So you can see really significant progress on that adoption curve, like, oh, we still have a lot more of that adoption curve to go. We also shared previously that our advertisers that have adopted our Conversions API and our privacy-safe measurement as a cohort tend to be 30%+ growers. Those that have not yet adopted tend to be mid-single digit decliners. And that's sort of a we're sort of swapping out the engines on the plane mid-flight. But what we're swapping to, we see that pattern holding consistently. Every time we get that adoption, we flip them into significant growers. So that's been really good. The other thing I'd say in terms of highlights on what's going really well and just increases our conviction as we look forward is that we're shifting the budgets.
Which budget do we get from the advertisers, right? And a couple of years ago, we mostly sat in sort of experimental budgets and social budgets that were not just awareness but tend to be sort of capped and are the first things you cut if you're an advertiser under pressure. Well, we're now moving into performance budgets, and we're doing really well there. In fact, when we look at our potential, you'd say, oh, well, Pinterest is 1% of the overall market. And can you be relevant at 1%?
Well, if you look at the largest, most sophisticated advertisers, particularly within retail and shopping, where we've talked about that's where we've had the most progress, and if you can make them happy, you can make anybody happy, in those places, we're seeing that we're getting north of for several of them, we're getting north of 5% of total ad budget. And then if you decompose that to the performance and digital side of the budget, that would mean that we're getting well into double digits on that. And so are we demonstrating that we can be a real player in performance? I think we have made a very clear case for that. And now we have to just keep driving through the adoption curve. So that's on the what's going well. On the, hey, what's a challenge? The adoption curve is real work, right?
I said we went from 2- 13- 23. Well, we want to get to 100. But there's real wood to chop there, right? The industry is going through a big shift to privacy-safe measurement. Advertisers are overwhelmed. A lot of them don't control their own tech stack. And they're scrambling to figure out, what does a cookieless world mean to me? And what do I do? And everybody's coming to me with different stuff. And what should I implement first? And so that's a lot of execution. And so we are laser-focused on that. And you've seen us make great progress on it. But we continue to be very focused on that.
Got it. OK. Let me ask one about that execution and sort of the macro backdrop that you're executing in. One of your larger competitors has been really increasing the amount of impressions across short-form video. And they have a lot of advertisers plugged in. There are other smaller platforms also coming along to advertisers saying, test our API, test our API. How do those two dynamics at a macro level impact your go-to-market and how quickly advertisers scale? And what is sort of the key pitch you make to the advertisers who have not adopted the API as to why they should choose you versus others?
Yeah. I mean, let's be very clear, the reality, if you're an advertiser, there are two platforms that are going to make up the vast majority of your budget, right? So if you're an advertiser, of course, you're going to go adopt latest tools from Google and Meta first, right? They're going to make up the dominant portion of your ad budget. And the case that we're making is, while of course you're going to do those first, we should absolutely be the very next in line. And the case we make for that is basically what I was just taking you through is, like, look at the largest, most sophisticated advertisers. And look at the ads running on our platform. You know those advertisers don't run ads in places that they don't see performance, right?
When you talk about the Amazons and Walmarts and Wayfairs and Targets of the world, if you're not delivering performance, they're not going to advertise with you. Well, anybody that's using our product can see, if you're doing anything with commercial intent, you're going to be seeing really great products and ads from those largest, most sophisticated advertisers. Well, those are sort of the lead buffaloes that the herd follows, right? And when we can say, oh, here's the kind of performance you can get. And by the way, we talked about Direct Links, where our earlier tools required the advertiser to do work. Like mobile deep linking, the advertiser had to go do work to implement. Conversions APIs had to do work to go implement. Well, we said, oh, well, Direct Links? We'll deliver that without them having to do any work.
Q4 this year, we shared this at the earnings call, we doubled the number of clicks we sent to advertisers year-over-year. Now there's a separation between value creation and value capture. That's value creation. So now when our reps are walking in, they're saying, OK, you may not have implemented our privacy-safe measurement yet. You may not be seeing this yet. But we doubled the traffic we sent to you year-over-year. Don't you want to see that? Don't you want to lean into that? That is one hell of a foot in the door for a sales pitch. At the same time, they're also saying, like, oh, well, I got to make sure I solve for the 80% of my budget that goes through two platforms first. And then where do I come in behind that? So we still have to work through that.
That's why there's still an adoption curve to manage through. But when you can walk in the door and say, we doubled the number of clicks we sent you for the same money. And we're able to drive performance for the largest, most sophisticated advertisers, the ones that are the winners in retail, don't you want to play there too? And by the way, we're not just talking about 1% of your budget. We can perform at 5%+ of your total budget. And we're doing that with some of these largest, most sophisticated players. That's a really great in with those advertisers. But there's still a lot of execution to do too.
Got it. OK. That's helpful. Let me ask you a couple about engagement. You talked earlier about sort of some of the strength of engagement and metrics that you've seen even beyond the reported user growth. So maybe let's just focus a little bit on U.S. and Canada, where I think the U.S.-Canada users sort of grew 1% or 2% most recently. Is that a relevant metric when you sort of think about KPIs that matter? And if it's not, what's another engagement metric that you're focused on just to ensure you can get enough potential engagement to monetize for these advertisers?
Yeah. Yeah. It's a great question. So I try to be pretty consistent. Try to keep a say-do ratio of 1. So go back to my first earnings call. I said, particularly in our mature markets, we're not going to chase MAU as a vanity metric. We've seen in our mature markets most of the users we need to see, right? Like approaching 100 million MAU in U.S. and Canada, those are most of the shoppers, right? There's some other ones you want to get. We wanted to get relevant with Gen Z, things like that. But I said then, in our mature markets, it's really about depth of engagement per user. We got to get from episodic usage to much more recurring usage in those mature markets.
And so if you look at what we've been doing, we have across our user base, we just put up our best user growth quarter since the height of the pandemic when people were locked in their homes and couldn't go anywhere. So we are finding our best product-market fit in years. I talked about how we have aged down with Gen Z now being our largest, fastest-growing demographic. But what you often see, and I've seen this many, many times through the course of maturing products that I've built and been a part of, is that those later cohorts are you get your power users first. Later cohorts are sort of like peripheral users. And so it actually erodes it typically. It erodes engagement per user because you're getting more fringe sort of users. We have consistently shown that there's user growth, MAU growth.
Our mobile app has grown faster than MAUs, which means more people are coming to us directly. So we're less SEO dependent. More than 80% of our users and revenue come directly to our app. Our basket of engagement metrics grows faster than users. And our ad impressions grow faster than engagement. And that one, by the way, I've talked about how when the user is in a commercial context and they have commercial intent, ads can be great content for the user. We are proving that out because even as we have taken ad impressions up significantly faster than user growth, engagement's actually deepening. That demonstrates that we are making the ads great content, satisfying that commercial intent. And so back to your question on MAU growth, it's like, yeah. We just put up our best user growth quarter since the height of the pandemic.
Our product-market fit is the best it's been in many, many years. As I shared at Investor Day, those new cohorts are roughly twice as engaged. The engagement per user and the fact that the ad impressions are able to scale significantly while engagement is deepening, those are the things that are the most exciting to me by far.
Got it. OK. And as part of that engagement per user, maybe talk to us about sort of what are the behaviors people are doing more of now than they were two years ago? And how prominent is search? How do we think about search use and then the eventual monetization of that?
Yeah. Great questions. So a couple of years ago, one of the questions was, like, well, hey, people browse a lot of stuff on Pinterest. Are they going to do anything? Well, part of what I saw from the outside is, like, they just hadn't been given the option. In fact, it was amazing. I've sort of given this metaphor before of, like, Pinterest has solved digital window shopping, but all the stores were closed. And so the one part of that was, like, oh, well, are they going to go in the stores if you open the doors? The other side was, like, how good was the window shopping? Well, it's so damn good that people keep coming back even though all the stores were closed. As we have opened the stores, we doubled the number of clicks we sent to advertisers year on year.
Our shopping ads continue to grow at 50%+ year-over-year. So that question of, like, will users take action on Pinterest, we have answered that emphatically. As I said, we are still relatively early in our journey on that. But we have answered the question pretty emphatically. So in terms of, like, what actions in that basket of engagement, what are the actions that we see versus a couple of years ago? A couple of years ago, the platform was optimizing for views and not optimizing as much for other actions like clicks, conversions. Then another one I give is curation, which I think was a severely underappreciated superpower of Pinterest. From the outside looking in, I saw that as, like, oh my god. That is some of the richest signal in the market.
In my past role, we didn't have that because it was, what was the user planning to do? Most of these shopping journeys are multi-day, multi-week, multi-month, depending on the complexity of the purchase. And getting that signal of not just what somebody clicked on, but what were they planning to do days, weeks, months before they did it? Those are magic moments for the advertiser where, like, the user has clear intent but hasn't decided what to buy yet. Like, oh my gosh. That's the perfect moment for an advertiser. So as we've leaned into that curation behavior, not only is that really great for how we can engage with advertisers, it's super rich signal that feeds into our relevancy. So coming back to your question on search, so on search, we've talked about this publicly quite a bit, that there's sort of three primary surfaces in our app.
There's home feed, search, and related items. It splits roughly a third, a third, a third across those. You could think of search and related items both as elements of search, right? Because related items are sort of a variation of a visual search. You found an item, and then it's, well, hey, I want more things like this. So it's a visual form of searching. I shared previously that as we had leaned into large language models, next-generation AI capabilities, that we were seeing models 100 times larger. As we are now training them on these intense signals, we were getting much better relevancy. I'd shared previously approximately a 10 percentage point lift in relevancy. There is a compounding effect of these things, right? Not only does the AI keep advancing, we are spinning the flywheel faster and faster.
As we lean into those curation behaviors and train our models on those curation behaviors, we just keep getting better and better and better at relevancy. So to give you a tangible example of this, our latest foundational models on computer vision that we are using for some of our biggest categories, like home goods and some of our core shopping use cases, we saw a 25% lift in the relevancy from those latest models. And it's, again, the AI keeps getting better. But the AI is only as good as the signal of how much it's acting. And that's where, from the outside in, when I was coming in, I was like, oh my gosh. This curation behavior is just like an amazingly rich signal that Pinterest wasn't taking advantage of. Well, not only are we, like, feeding that into the models, it makes the relevancy better.
We're designing the product to capture more of that curation. That flywheel just keeps spinning more and more and more. That is giving us, to the question on search, we are seeing that users are searching more on the platform as a result. We also talked about Guided Search. Investor Day, we previewed it a bit. We talked about it on the last call. Guided Search is we are a visual search and discovery platform. One of the things that happens so often with search is that people don't have the words to express the thing they're looking for, right? If they had the words to express it, you get this, like, really long you even hear this with Gen AI. There are prompt engineers.
It's like, oh, well, if you need a prompt engineer, for the average person, it's like, this is sort of the core issue. They don't have the words to describe the thing they're looking for. Well, Guided Search is a way of enhancing, with generative AI, further enhancing how people would explore a topic where they might not even know that, oh, well, that beach house I was designing or the theme for a beach house is actually Cape Dutch/Moroccan. I didn't even know that. I just saw a picture and I thought it looked cool. And with Guided Search, we start to give people the language for that and help them refine in on these things. But as we have been rolling out our own Guided Search experience, we're seeing that that, paired with relevancy, users are searching more and more on the platform.
So we feel really good about the progress there, which, again, just aligns with we're very distinct and separate from entertainment-based social media that has users in a lean-back entertainment mode. We have users in a lean-forward intent-based mode. And as we're doing those things, we're getting more and more of that repeat usage of the depth of engagement that I was talking about.
Got it. OK. Let me ask another one on Gen AI. I'm going to sort of pair it with margins. As Julia announced, let's pay attention. You talked about the margin guide and the margins you've delivered so far. Low 30s Adjusted EBITDA margins. They were 37%, 24%, and 37% the last two quarters. So you've done a very good job on the margin side. What are examples where you can use or already are using generative AI internally to drive more efficiencies? And how does that sort of impact your expectations around hiring the next couple of years and the overall OpEx investment?
Yeah. Great question. So AI is absolutely a core competency for us, right? And we've talked about that in terms of what we're doing in the product and just tremendous yield there, as I was just describing. We are looking at efficiency as well, though. I think one of the biggest highlights, and I think this really stands in stark contrast with a lot of others, that the question in many other places is, like, "oh my gosh, isn't this generative AI margin headwind for you," right? And so there's two parts. One, we're aligning generative AI with use cases that really pay out well. So we may spend more on a GPU than a CPU.
But when we can go drive up ad relevancy and shift from impressions to conversions, the conversion pays out a multiple of the impression, we're just aligning the use of the GPUs and the generative AI capabilities with high ROI use cases. And we can do that because we have real commercial intent from our users versus if we were doing huge view time on video, you have to have a lot of view time to get to one impression. And you need a hell of a lot of impressions to get to one conversion. And so there's some natural efficiency in the platform. But then back to the internal side of it, yeah.
We're seeing that we've talked about some of this, that as we look at coding use cases and things like that, we're seeing that for a lot of the drudgery of coding, writing your unit tests and your documentation and all these things, it can be really helpful for that. And it can perform really well for early coders to help them get more productive more quickly. And so we're seeing really nice efficiencies there. But it's almost across every function of our business that we're seeing really good benefit across that. We have more than 50% of our engineers are using Copilot-type tools for their coding, right? And we're seeing real productivity improvement. But in terms of what it means for hiring, just as on the infrastructure side, our margin expansion, infrastructure was a real highlight in that, that we actually reduced infra spend.
Even as we increased usage, increased our revenue, we reduced infrastructure spend. We think there's some more juice left in that orange as well. We grew headcount at 1%. I actually clamped down on headcount pretty much as soon as I came into the business. That was before it was in vogue, before everybody was sort of focused on efficiency coming in. I'd seen, like, oh, wow. Margin's really eroded. We're going to get really focused on how we drive operational rigor in the business. Back to say-do ratio of 1, in those early calls, I said I'm a big believer in theory of constraints, that oftentimes you don't need more engineers doesn't necessarily equal a better outcome. Even as we've held headcount to 1% growth, we're finding our best product-market fit in years.
We've accelerated product velocity because we've been much more focused on what matters most to users. As we've focused and channeled that energy to the right things, we're getting great yield. People are excited because, oh, yeah. People like to build. They like to build stuff that people use. They want to build things that have an impact versus shipping a lot of features. Pinterest had become a bit of a feature factory where it's like, oh, there's all these features. Like, well, that's nice. But if they don't have impact, it's not just from an investor perspective. You'd say, well, that's not that exciting. If it doesn't have an impact, well, if you're the person building it, you don't want to build stuff that goes unused. You want to build stuff that has impact.
And so there's another sort of flywheel around that as well, that even as we've had really good discipline on headcount, people are seeing, like, they're building stuff that matters. And the velocity has increased. And the product-market fit has increased. And so that's been a real highlight for us.
Got it. OK. So it sounds pretty efficient on the engineering side. And then the sales side, we have the third-party partnerships I want to talk to you about. So first, before we sort of get into the specifics of Amazon versus the Google partnership, one of the most common questions that myself and my team continue to get from investors is how to think about the contribution from the third-party partnerships in that multi-year guidance you laid out at Analyst Day. So maybe just remind us, what is built into that mid to high teens growth and sort of the potential for upside to it vis-à-vis third-party partnerships?
Yeah. So I've been very consistent on this as well to say that we think about third-party partnerships as a way to round out gaps in our auction. And by the way, we're not unique in that, right? Even the largest platforms when you think about Performance Max and things like that, even the largest platforms in the world use third parties to round out gaps in the auction. Now, as a smaller platform, we had bigger gaps in the auction. And so when we talk about improving that actionability, that was our first focus is we want to go drive more relevant ads? We'd better have the right demand. And it better be good shoppable content. So that was our focus, is how do we bring in third-party demand that can really round out those gaps in the auction?
But it's important to remember, it's going into one auction. And so one of the reasons that we consistently said we're not going to break that out separately because it's going into one auction. It's not like it's some totally different product line or something like that. It's all going into one auction. And in the same way that we wouldn't break out one retailer, it's one auction. And we're seeing we said this on the earnings call. And it's consistent with the timeline I laid out from the very beginning, that I said from the beginning that it was the first time that we were opening up to third-party demand. It was going to take us multiple quarters to do that. That's pretty meaningful work to do that, particularly when you focus on relevance, right? You're just trying to do display ads that aren't relevant to the user.
That's easy. But it's not great for the user. We're actually doing ads that help engagement versus take away from it. We said that that would be multi-quarter and that we'd start to have more meaningful contribution in early 2024. Well, that's exactly what we said on our call, that we are now seeing as we get in as we're coming into Q1 that it is a more meaningful contributor. But we're not breaking it out separate. But we see it as a meaningful contributor. It's not dominating our growth. I've heard the question come in another form of, like, oh, well, does this basically make you a publisher versus an ad platform? Well, in no way do we expect that 3P is going to dominate our platform. It's filling in gaps in the platform, right? And even as we had larger gaps than others, it's not dominating our growth.
It's a meaningful contributor. But it's not dominating. And even as we are now expanding the aperture because I also said from the beginning that we would have multiple partners because we have different gaps in the auction. The first thing to address was shoppability. Amazon's been a fantastic partner on 1P and 3P around bringing great shoppable content into the platform. By the way, other large retailers have been fantastic as well. Walmart, Target, others, Wayfair are bringing great shoppable content into the platform. And now we see more and more retailers piling in. That's been great. We announced the partnership with Google focused on international, where we have roughly 80% of our users outside the U.S. but 20%-ish of our revenue. For larger platforms, you'd see roughly 50/50 in terms of U.S. versus international on revenue.
Google's got fantastic reach in those international markets that were previously unmonetized for us. So that's another gap in the auction. We're working with them on that. But we fully expect that first party continues to be the largest driver in our business. And third party helps to round out gaps. We also talked about in our sales approach that we'll be first party in the largest markets, U.S. and international. And we are that today. We're working with resellers. That is another sort of a multi-pronged approach, right? So first party in the largest, most mature markets, resellers that help us get into sort of the next stage of markets. And then 3P is a nice enhancement to that as well. So you see us continuing to execute on that multi-pronged strategy. And across each of those, we're making great progress.
In no way do we expect that our future is just that of a publisher and not somebody that is predominantly selling ads directly. Absolutely. That is what we will do predominantly. These third-party partnerships are meaningful contributors. They're bringing really great ad demand into the platform that not only contributes to revenue but actually contributes to user engagement as it helps us solve actionability.
Yep. OK. As I know, we and investors will continue to do a lot of ad checks and agency checks about these third-party partnerships. There's always a lot of math thrown around about how big they can be. And sometimes I worry the math is candidly just too simple. It misses the point of incrementality and sort of the auction density. So if you're looking at it externally, what are the hurdles that we should be mindful of that are going to sort of make or break how much incremental revenue you can get from these partnerships just so that there isn't this, oh, ARPU is this. And you can take ARPU up. What are sort of the key executional hurdles that are going to make or break Amazon and Google being bigger or smaller?
Well, I mean, one of the biggest ones is just, like, is it relevant to ad demand, right? And so I've talked about that previously, that we're seeing good relevant ad demand. With the Amazon partnership, Google, we're just getting going. But obviously, they've got fantastic reach across those unmonetized markets for us. So we have good expectations of what we can do there. And so I think that's the biggest one. At the same time, these things do require tuning, right? And so it takes a while to tune. And we will be tuning and scaling in perpetuity, right? So sometimes you get the question of, like, hey, when are you done tuning? Well, never. I've said the same thing in my old job too, right? Like, no. You're constantly tuning. But I think we are a long way from having seen the ceiling on these.
We are still early innings on it. But we have more than proven out the theory of the case. And so that tuning and scaling, consistent with what we said, we're right on track with what we said in the implementation. We're seeing 3P now become a more meaningful contributor to our growth. It's not dominating our growth. But it's a meaningful contributor. And as we continue to tune and ramp with Amazon, bring Google in on international, we're not done, right? We'll have other types of gaps that we'll fill in as well. We'll continue those things as we go forward. I got this question a lot previously, like, hey, are you doing something bespoke? Or are you building it for multiple? It's like, no. We're building to ingest multiple. We've proven that out. But we feel good about the progress there.
On the Google international point, is there any sort of plumbing or strategic reason as to why Google is not in the U.S. yet? Or is this sort of a question of when, not whether? Why international? Why not Google?
There's no structural reason we can't do that. There's nothing that precludes us from doing that. We're just focused on, like, OK, what's our next biggest gap to fill in, right? A year ago, biggest gap was solving actionability and shoppability on the platform. Well, Amazon has a fantastic shopping and buying experience and really great shoppable ad demand. And so that was absolutely the right place to start. We feel good about the progress there. More to go. But feel good about the progress on that front. Next biggest gap, 80% of our users outside the U.S. but only 20% of our revenue. A lot of these markets are completely unmonetized. That's the next biggest gap. So we're really focused on the Google partnership in solving that gap. But there's nothing that precludes us from only doing that.
But I talked about earlier of, like, we've accelerated product velocity and product-market fit through clarity, focus, good execution. We're going to have clarity and focus and good execution on these things. And then over time, you can start to blend them even more. But for now, we're focused on what's our next biggest gap to go solve. And international clearly is a big one.
One of the interesting couple of slides you had at the Analyst Day, I thought, was when you sort of talked about the potential for even further automation in the advertising tools for advertisers. Google has PMax. Meta has Advantage+ . Those tools do seem to be quite effective. How far away are you from being able to really push those types of tools out to more advertisers to just make the whole process more automated?
Yeah. Great question. So we've made phenomenal progress on sort of AI-driven tools for advertisers. The most tangible example would be automated bidding, is now 85% of our revenue, right? And we've seen rapid adoption of that over the last couple of years. And I don't think it. I was at Google when PMax was being built. It was very close to that. And it eventually got a name. But there were lots of components of that that you're sort of building up all the components. And it's like, oh, yeah. Now we can call this a suite. And so similarly, that's what we're doing as well. All the components of this, we've been building up. And we're across most of the components. Each of them is maturing. Automated bidding, we're the furthest along. We're now going deeper into broader campaign management, things like that.
As we go forward, you'll get into things like dynamic creative. We talked about that a bit at Investor Day. Those are some of the things that are sort of still out in front of us. But where we think there's really good opportunity, right? And where we think we can shine. Like on the dynamic creative side, future state of the world for us, oh, you and I both love coffee, I'm assuming. I hope you love coffee. You and I both love coffee. We both want, like, a really nice Breville coffee maker. Well, but maybe you like Scandinavian minimalist design. I'm doing a Cape Dutch Moroccan-themed beach house. Well, when you see the coffee maker, you see it in a Scandinavian minimalist design background. I see it in a Cape Dutch Moroccan beach house.
And we both convert better on the same product because of that kind of dynamic creative. And so those are the kinds of things that we can do in the future, not just because of the same Gen AI that everybody else uses but because we have really unique signal. Because that level of knowledge of, like, what's your style versus my style, that's the kind of really unique signal that we get. So we continue to build out the suite. We see really great yield from those things. And one last point I'll make on this is we continue to round out the suite. It's another example of just compounding benefits, right? I talked about how advertisers have at least three of our lower funnel tools, how we're growing that, and how we move people from mid-single digit decliners to 30% growers.
Similarly, on AI-driven tools for the advertisers, as they adopt more components of this, compounding benefit to that as well in terms of how they shift spend to us and those kinds of things. So we feel good about the progress there. While we're not all the way to the place of, like, a PMax or Advantage+ , you're comparing to best in the industry. If you compare us to other sort of mid-size advertisers, I think that's one of the things that goes underappreciated is, like, we have really distanced ourselves from the pack there, right? We get this question of, like, oh, is it sort of the largest platforms and then everybody else? Well, that's one way to look at it. And there is an advantage of scale. Another way to look at it is, like, there's going to be those with commercial intent and everybody else.
I think us leaning into that in this curation behavior and the personalization of that is letting us really distance ourselves not just from other mid-size players but even from the largest players when we're competing for performance. I talked about the progress that we're making there and breaking into those performance budgets, getting north of 5% of total ad spend for these largest, most sophisticated advertisers, double-digit on their performance budgets. We are really proving our mettle there. Yes, the full automated suite, we continue to work on that and see great progress there too.
Great. Bill, thank you very much.
Thank you.
I'm going to get a coffee. I prefer it at your beach house. But Bill, thank you so much.
All right.
Let's see what happens.
All right. Thank you.