Snap Inc.

Q4 2025 Earnings Call

will be partially offset by reduced spending on community growth marketing as we adjust these investments to better reflect the long-term monetization potential of each geography. As a result, we estimate that full-year adjusted operating expenses will be approximately $3 billion. For SVC and related expenses, we estimate approximately 1.2 billion in 2026. For Q1 specifically, our guidance range for revenue is 1.5 to 1.53 billion. Our Q1 revenue guidance range excludes any potential revenue from the perplexity integration as we have yet to mutually agree on a path to a broader rollout. Given this revenue range and our investment plans for the year ahead, We estimate that adjusted EBITDA will be between 170 million and 190 million in Q1. As we begin 2026, we are excited to execute on our pivot towards profitable growth and to make incremental progress toward our medium-term goal of delivering meaningful net income profitability. The impacts of this strategic direction are already evident in our Q4 results, and we are incredibly proud of the work our team is doing to build on this momentum in Q1. Thank you for joining our call today, and we will now take your questions. Thank you.

We will now begin the question and answer session. To ask a question, you may press star, then 1 on your touchtone phone. If you are using a speakerphone, please pick up your handset before pressing the keys. To withdraw your question, please press star, then 2. In the interest of time, we ask that you please limit yourself to one question. After your initial question is asked, your line will be muted.

At this time, we will pause momentarily to assemble our roster.

The first question comes from Eric Sheridan with Goldman Sachs. You may proceed. Thanks so much for taking the question, and thanks for all the depth of details in the updates on the quarterly results. I want to talk about one of the forward initiatives, Evan. You know, with SNAP specs as one of the key priorities in the next one to two years, can you just go a little bit deeper into what you've built on the platform and the application and use case side and how you think it feeds into where you want to take the hardware side of the business when we think about the next 12 to 18 months and how this fits into your broader strategic priorities for the company and more particularly for spatial computing longer term. Thanks so much. Eric, thanks so much for the question. We're super excited about what's ahead this year with the launch of specs and obviously graduating from the R&D phase of specs to, you know, broader consumer adoption. And in preparation of that, you know, we've been working on several prior versions of specs, including most recently the version released in 2024 to developers, you know, who can subscribe to to specs and start building lens experiences.

We've seen some people build really spectacular things, whether it's utilities or new educational tools, for example, like an at-home chemistry lab. You can have an augmented reality to even some of the more interesting work we've been doing with the browser and the ability to stream video on a virtual screen grounded in the real world through your glasses. So it's been really exciting to see all the new use cases that developers are building for specs with the current version released back in 2024. And those will be able to run on the forthcoming version of specs released later this year. So I think we'll be able to launch with a really wide variety of compelling experiences, which I think is so important for the early success of a product like this. And We're just really focused on getting the hands of early adopters. We're so fortunate to have this passionate base of developers, you know, hundreds of thousands of developers who have used Lens Studio to build lenses. And I think they're really excited about this forthcoming product. So really trying to engage them and early adopters with specs later this year. is super exciting.

And, you know, I think as we look out, you know, to future generations of the product through the end of this decade, you know, we've got a really clear path here to, you know, lightweight, affordable, and incredibly powerful glasses that can deliver immersive experiences in the real world. Great. Thank you.

Thank you.

The next question comes from Ross Sandler with Barclays. You may proceed. end of the range and it also yep can you hear me we can now yeah go ahead we can now go okay sorry okay the one cue guide uh assumes a pickup and growth at the high end uh and you guys mentioned that there's no perplexity in there could you just talk about what's driving that between dr and brand and how you're kind of expecting trends in 2026 in the ad business to play out. Thank you. Hey, thanks for the question. On the ad side, the biggest focus is continuing to generate additional demand by demonstrating the strong performance of the ad platform. So, you know, at the top of that, we're seeing really strong growth in active advertisers. They were up 28% year over year in Q4. As we continue to invest and scale our S&B go-to-market operations, And that's something you're going to see us build on into 2026. That's part of the investment plan for the year ahead is to continue to scale that out so that we can build on the momentum we have there.

We've seen especially strong growth in the medium customer segment globally, with medium customers in North America in particular being the largest contributor to absolute dollar growth there, which is good. So that's the kind of momentum we want to build on in 2026. We do continue to face some headwinds in the North America large customer business, but there are some bright spots there. including U.S. LCS financial services, vertical, as well as autos. You know, we have new leadership in place over the North America LCS segment. We've got new products to connect brands with Snapchatters, including sponsored snaps and promoted places to build with there, and smart campaign solutions to make it easier for advertisers to leverage the full set of Snapchat placements to make those connections easy and performant. So those will be big themes that we'll be building on in 26 as well. In terms of the guide for Q1, you know, the macro operating environment has thus far remained relatively stable compared to what we saw in Q4. There's a lot of quarter left to go in Q1, of course, but our guidance range is built on the assumption that the macro environment continues to be stable.

I hope that extra color helps a little bit.

Thank you.

The next question comes from Rich Greensville with LightShed Partners. You may proceed. a couple of questions first you know the subscription side which i know evan if i go back to your um your letter a while ago you sort of marked the importance of subscription it seemed like it really accelerated this quarter and i'm curious are you marketing it differently are there new features that you added i know you're you've talked about sort of charging for memories and other things that will add to this but just in terms of what happened in q4 it'd be great to better understand what's happening inside of that snap and then The other thing, I think two years ago, Evan, you got on this earnings call and you talked about the fact that you were sort of refocusing user growth efforts from Android developing markets to the bigger markets like the U.S. where the meat of your monetization was.

And, you know, if I look at sort of where U.S. users or North American users have fallen to at $94 million, Do you need to put even more effort into those efforts to sort of drive U.S. users or North American users? Just what's happening in the North American user market would be great to just better understand, given your focus there. Thanks. Yeah, thanks, Rich. We're definitely excited about what we're seeing on the subscriber side of the business. You know, certainly memory storage plans were a big driver of the subscriber growth. growth that we've seen recently, and also, you know, have helped improve retention rates overall. So that definitely, you know, has been really helpful to the subscription business. And we've got some other great, you know, features on deck coming up this year for the direct pay segment of our business. So really excited about that overall, and I think really helps support our efforts to diversify our revenue in addition to the small and medium customer growth that Derek mentioned. So overall, really excited about the progress on subscriptions and the diversification of our revenue.

You know, as it pertains to user growth, I think, you know, if you take a step back and look at the growth overall of the platform, you know, monthly active users now 946 million. So we're pretty close to our goal of a billion monthly active users. And I think, as you know, over the past three years, our community growth has really outpaced our revenue growth and ARPU has actually declined while we've simultaneously increased the cost to serve, which has put downward pressure on our margins. So as we look at this crucible moment and the pivot to profitability, we have immense daily reach and engagement in many of the most valuable advertising markets, including in North America. And we think we can strike a much better balance between pursuing community growth and also growing average revenue per user. So In addition to that, obviously, we're working through some of the regulatory landscape and some of the shifting user engagement patterns as we focus on organic growth.

But I think taking that all in totality, we've made some choices to reduce community growth marketing spend, to adjust the cost to serve, and to roll out additional paid features like the memory storage plans that we just discussed. And all those can cause headwinds to user engagement. So those changes actually free up more resources to focus on our most valuable geographies. so that we can continue innovating and delivering great customer experiences, which we really believe is the most important driver of long-term growth.

Thank you.

The next question comes from Dan Salmon with New Street Research. You may proceed. Great. Good afternoon, everyone. Evan, I wanted to just talk a little bit more about, as you called it, the sort of litigation or regulatory risk. caused by changes in age verification policies, sort of broader teen smartphone and social media restrictions. You obviously commented on the actions that you took in Australia following the ban going into place there. But what I'm particularly interested to hear a little bit more about is the potential for those types of actions to impact North America. Obviously, a $4 million step down in the DAU this quarter. I'm curious just maybe to unpack a little bit of what drove that more and what the outlook could be there during the year based on some of those litigation risks or regulatory risks you mentioned. Thanks. We're certainly aware of some pending legislation. Obviously, there's quite a bit working its way to the court system right now that would further restrict the use of Snapchat for our community. I think as we look at, for example, global ad revenue from impressions served to users under the age of 18, that revenue is not material.

So I think, you know, looking at sort of the revenue-generating potential of business looking forward, you know, we're not overly concerned about the changing regulatory environment. I will say, you know, one of the things that's very interesting is that if you look at the research studies that look at Snapchat specifically as separate from some of the studies that look at social media in totality. You know, I think what we continue to see, which, you know, makes us proud of the service we've developed, is that Snapchat actually has a positive impact on people's well-being and people's friendships. And that's actually in contrast to other services that don't necessarily have that positive impact. But I think we have had quite a bit of trouble as we look at the regulators explaining how different Snapchat is because there is really this moment you know, where people are expressing concern about, you know, use of social media. So we have to continue making the case that, you know, Snapchat and its orientation around your close friends and your family can have a really positive impact.

I think that's backed up by the research, but certainly it's going to take time to prove that out, and especially as these regulations sort of work their way through the court system. Thank you. The following comes from Ken Goreski with Wells Fargo. You may proceed. Thanks. Maybe first I'll touch on specs. Could you talk about, maybe Evan, can you talk about the kind of synergy between specs and Snap services more broadly and the audience and kind of the developer base? And then talk about, you know, the right way to capitalize that entity. I mean, if there's... If you have confidence in the end product, how do you think about appropriately capitalizing that? Should it happen all within Snap? Should there be outside partners? And how do you accelerate kind of the development and the deployment of specs throughout the ecosystem? I'll stop there. Thank you.

Yeah, well, I think to just maybe take a step back on why we started working on specs in the first place, you know, when we invented Snap and we worked on things like ephemeral messaging or stories that put content in chronological order or even things like opening to the camera, our vision, our work was really designed to make, you know, computing or smartphones feel more human. And we think that's played a really important role in connecting people with their friends and their family. But we also saw a lot of limitations. you know, of the smartphone and of computers. And I think today people are spending something like seven hours a day in front of a screen. And so I think there is at this moment a real opportunity to change what the computer is instead of something that, you know, you're constantly operating using a keyboard and a mouse, something that, you know, now powered by AI can actually get work done for you. And so in that way, it's really a continuation of this vision to try to work to make computing a more human for folks. And so I think now that we are exiting the R&D phase of specs development, there's a couple important things.

One is developing a strong standalone brand. I think specs, the product itself, in many ways appeals to a different audience segment than the core Snapchat audience. And it's going to be really important for us to develop a standalone brand identity. for specs. And then I think longer term, you know, as we look at the rollout and broader deployment of specs, there may be opportunities to, you know, raise additional capital to accelerate, you know, balancing that obviously with our own sort of ownership interest and any potential dilution. So I think right now, you know, given that we're so close to launch, the key here is really just, you know, nailing the launch and making sure that we deliver an extraordinary product. And then, you know, I think we have a lot of flexibility to think about how we want to capitalize it

Thank you.

The next question comes from Justin Patterson with KeyBank. You may proceed. Great. Thank you. Good afternoon. I wanted to talk about agentic coding. We've seen more companies see meaningful improvements in engineering productivity from these tools. How is this being deployed at Snap today? And how should we think about potential benefits, whether it's product velocity, more engagement on the platform, more monetization opportunities, or expense efficiency? Thank you. Yeah, there's just so much opportunity here, obviously. You know, I think now something like 40% of new code at Snap is AI generated. We made a ton of headway with, you know, trust and safety and customer service. in terms of automating those workflows. I think there's a lot of opportunity for the sales workflow as well to empower our sales team, but also to automate quite a bit of that. So certainly we're seeing gains across the board in how we're operating our business today. I also think this can be a real accelerant for our own creativity.

I mean, one of the things we love to do is invent new services, and we've got a bunch of ideas for new apps, for example, that we could build using these AI tools and deploy very, very quickly, leveraging, of course, the distribution we have, our friend graph, some of the unique assets we have, like folks' memories, for example. So I think there's a lot of opportunity here for us to think about how we accelerate the growth of our business and actually develop new services that quickly using these tools. And I think in addition to that, we're just running as fast as we can to roll out new agents across the enterprise, new tools. And especially for a small team like the one we've got at Snap, this is just a massive force multiplier. And I think really will help accelerate a lot of the creative vision we have in terms of turning it into reality. Thank you.

Our last question comes from Benjamin Black with Deutsche Bank. You may proceed. Great. Thanks for taking my question. You know, can you talk about the decision to moderate infrastructure spending at a time when others are ramping spend to drive ad performance? Was there sort of slack in the system? Maybe just talk us through that decision. Thank you. Hey, it's a great question. Thanks for asking it. I think, you know, the first thing I would say just for context, the big driver in the ramp of infrastructure investment over the last couple of years has been a really significant growth in our ML and AI investment, and that was to both support the rebuild of the ad platform and the DR advertising business, and also to support the content business and ranking and personalization and all the work that we've done there. And I think I would say, first and foremost, we intend to continue to invest pretty heavily there. And so that's not an area, you know, focused for pulling back.

You know, as it pertains to infrastructure specifically, there are really two big catalysts where we see a lot of opportunity and are already making progress in terms of driving, you know, margin efficiency for the business and margin expansion. You know, the first there is just our investments in how we handle cost to serve. and getting that in a place where we're calibrating that better relative to the monetization potential of each of the markets in which we're operating. And there's a lot we can do to optimize that. And that's really about the theme that we've been talking about in terms of getting to profitable growth. And so translating that, you know, into the growth and infrastructure really being keyed in against the growth in monetization. The other real opportunity we see here is to take you know, some of the infrastructure things that are costs right now and turn them into revenue generating investments.

And so I think the recent launch of the memory storage plans is a great example of that, where we can take a cost and not only, you know, find ways to make it more efficient, but then also turn it into a revenue generating source of top line growth, which is going to help with even further margin expansion. So a lot of this is about efficiency. A lot of it is about being really sensible about our cost to serve relative monetization potential markets and then scaling efficiently. But those investments in AI and ML will continue to be really important to the performance of the business and both the ads and the content side. So hopefully that gives a little bit more context there. Thanks for asking. This concludes our question and answer session as well as SNAP Inc.' 's fourth quarter 2025 earnings conference call. Thank you for attending today's session. You may now disconnect.

Q3 2025 Earnings Call

ons towards sponsored staffs and spotlight helped to reduce this to 18% in Q3, and we anticipate being in the 18% to 19% range in Q4. For adjusted operating expenses, we provided full-year guidance of $2.7 to $2.75 billion, which we reduced to $2.65 billion to $2.7 billion earlier this year, And we currently estimate we will end the full year nearer the low end of this reduced range. For SBC and related expenses, we got it for a range of 1.13 billion to 1.2 billion for the full year. We reduced this to 1.1 to 1.13 billion earlier this year, and now estimate we will come within a further reduced range of 1.08 to 1.1 billion. Given the revenue range above, and the progress we have made to optimize our cost structure, we estimate that adjusted EBITDA will be between $280 million and $310 million in Q4. Given the strength of our balance sheet, our progress towards sustained free cash flow generation, and our desire to opportunistically manage our share count for the benefit of our long-term shareholders, we have authorized a new share repurchase program in the amount of $500 million.

As we look to close out 2025, We are excited by the opportunities ahead of us to accelerate top line growth, further diversify our revenue sources, and make meaningful progress toward profitability in the year ahead. Thank you for joining our call today, and we will now take your questions. Thank you. We will now begin the question and answer session. To ask a question, you may press star, then 1 on your touchtone phone. If you are using a speakerphone, please pick up your handset before pressing the keys. To withdraw your question, please press star, then two. In the interest of time, we ask that you please limit yourself to one question. After your initial question is asked, your line will be muted.

At this time, we will pause momentarily to assemble our roster.

The first question comes from Rich Greensfield with LightShade Partners. You may proceed. Hi. Thanks for taking the questions. I've got a couple. On the perplexity partnership, which is really interesting that you're going to add it on to Snap AI, is the cash stock split already determined, Evan, or could it actually change based on factors that you can help us understand? And you talk about monetization for the partnership starting in 2026. Does Snap ad sales, like will your ad sales team be selling ad units that appear in perplexity? Or just help us understand what monetization could look like inside this perplexity bot that's going to live inside of Snap. And then just a question for Derek. On a two-year stock basis, it looks like cost of revenue really came down. You talked about a shift to spotlight and sponsored Snaps. Should we presume that the reason why we're seeing that leverage in cost of revenue is because you're not paying out to content owners the way you do in Discover for those ad units? Just would love to understand those two main things. Thanks. Hey Rich, thanks so much for the question.

We're really excited about the Perplexity partnership and I think it sort of underscores Snapchat's role as a messaging service and how valuable that is in the age of AI, you know, especially because Snapchat engagement is built around real relationships between friends and family, but also because conversational assistance is very quickly becoming the primary way that people are choosing to interact with information. on the internet. So I think, you know, we have a really unique opportunity ahead to help distribute AI agents through our chat interface and launching with Perplexity next year to bring their answer engine to Snapchat, you know, is really in the default placement in our chat inbox is going to be really valuable to our community and hopefully very valuable to Perplexity and to Snap as well. And I think Perplexity's focus on trusted and verifiable sources really aligns with our values and makes them a good fit for our community. I think To answer

your question from a monetization perspective, we don't expect to recognize any of the $400 million until we begin to roll out the integration, likely towards the beginning of next year. And Perplexity will control the responses from their chatbot inside of Snapchat. So we won't be selling advertising against the Perplexity responses. But I do believe that the placement will help Perplexity drive additional subscribers, which I think is something that will be valuable to their business i think just looking ahead one of the things that's really exciting is the opportunity to expand to more partnerships and advertisers are very focused on leveraging sponsored snaps to distribute uh conversational commerce experiences uh with their brands so that's something we'll be experimenting with uh you know and you know as we uh as we kick off next year thanks hey it's derek speaking i can take the cost element of that question You know, as Evan noted in his letter earlier this fall, we see a lot of opportunity to expand our gross margins, and we're working across a number of different fronts to achieve this, including by improving the top line growth as well as becoming more efficient on cost of sales. So on the revenue front, we're broadly taking steps to better monetize our core product value. So we see sponsored snaps and promoted places were first steps on that journey. The ongoing growth of Snapchat Plus, the introduction of Lens Plus, and now The recent announcement and testing of memory storage plans are all examples, with the latter being a great example of an area where we can flip a cost structure into a revenue generating business line. The perplexity deal is yet another example of a new line of revenue generation that helps expand the margins also. On the cost side, we see several dimensions to this, including work to optimize our content programs, recalibrating our investments in community growth and the cost to serve our community to better match the long term financial potential of each market. In Q3 specifically, we're seeing the benefit of a mix shift in where impressions are being delivered, you know, in particular to sponsored snaps and to a certain extent to spotlight. And as you've noted, these surfaces have higher margins, and this contributed directly to gross margin improvement of 55% in Q3, up from 52% in the prior quarter and 54% in the prior year. So lots of work to do there, but we're excited about the progress when we saw there in Q3. Thanks for clarifying, Derek. Thanks. Thank you. Next question comes from Mark Schmuck with Alliance Bernstein. You may proceed. Yes, thanks for taking the questions. You know, Evan, just to follow up on that last answer, kind of beyond perplexity, you know, how do you see Snaps' role evolving here as kind of this distribution channel? Sounds like there may be something about kind of brand messaging integrations, but, you know, could we potentially see the ad stack open up as well? And then, Derek, kind of on the commentary around the Q4 engagement headwinds, you know, If we try to compartmentalize that, is the bulk of that kind of like one time in nature as we kind of think about some of these regulatory type headwinds and then we kind of rebuild the ramp from there? Is that the right way to think about it? Thank you. Hey, thanks so much for the question. I think, you know, as it pertains to opening up the platform further, what we're seeing is a lot of our advertising clients are investing a lot in these, you know, conversational experiences, whether they're educational or really designed to improve consideration or folks who are going and developing full-fledged commerce experiences inside their own chatbots. But despite all this investment in building out that customer experience, folks are struggling to find distribution channels for those experiences. And so while there's a lot of development of AI agents right now, I think we're very quickly seeing people shift their focus to trying to develop more distribution. So I think Given Snap's primary engagement around messaging, there's a real opportunity to open up our chat inbox and chat interface to more of these agents and to really to distribute them through our sponsored Snap's products. So that's an area of investment for us. The work we're doing to support perplexity and the development of our APIs there will also support other partners over time. And it's certainly something we're excited about. I think it's also a really compelling customer experience, given what we're seeing in the way people are shifting their behavior patterns to, you know, engage with these chatbots. So definitely as a chat service, we're very excited about the, you know, evolution in the customer expectation there. You know, I think as it pertains to the overall DAU growth and our efforts, we've been doing a lot to overcome ongoing engagement headwinds to DAU primarily by introducing new conversation starters. So if you think historically on Snapchat, a lot of conversations have been started by folks replying to friends' stories. As we've seen engagement shift from things like friends' stories to content posted on Spotlight, for example, we have to migrate that friends' story reply behavior to things like sharing Spotlight videos or reposting Spotlight videos or playing games with friends. We've got some forthcoming product initiatives as well that are oriented around new ways to start conversations and spark conversations with friends. I would say big picture though, in terms of our growth in daily active users, I outlined in my letter, I think that was released back in September, this crucible moment for Snap and really a pivot to more profitable growth during this period. know we tried to provide a few examples in the in the earnings uh release but we're experimenting with things like changing the way we do pre-fetching and caching and certain uh geographies for our content business you know changing the candidate size uh you know number of candidates essentially in our ranking retrieval systems and certain geographies and really trying to line up our infrastructure and marketing investments against the the geographies where we see the la you know the the largest uh long-term monetization potential. So I do think there will be trade-offs there in terms of engagement. But ultimately, as we focus on more profitable growth, I think those are trade-offs that we're going to want to accept. And then we're also, of course, investing in things like memory storage plans or LensPlus. I think those are things that reflect the real in terms of large-scale cloud storage or new AI tools in LensPlus, but those also add some friction to the user experience. And then I think perhaps most importantly, we're going to be proactive. We're going to get on the front foot when it comes to rolling out age assurance. There are new age assurance signals that Apple is providing us. This quarter, we're going to use those signals to detect underage users. I think Google is rolling out a solution as well, perhaps at the beginning of next year. And so as we roll out those age assurance signals, that may have an impact on daily active users as well. But we think that's the right thing to do. It's important for maintaining trust with our community and, of course, as well with regulators. But that could be a headwind to growth as well. Thank you. As a quick reminder, if you'd like to ask a question, please press star, then one on your touchstone keypad.

The next question comes from Doug Anmuth with J.P. Morgan. You may proceed. Hi, this is Maggie on for Doug. Thanks for taking the question. We can tell that mid-sized advertisers are clearly a focus for SNAP. Could you just expand a bit more on your go-to-market efforts and product roadmap to unlock greater spend from this segment? Thanks. Yeah, we're so excited about the growth we've been seeing with our small and medium-sized customers. We've obviously got very strong product market fit with our app product, lead gen, of course, you know, our web direct response product as well. A lot of what we've been focused on from a product perspective, things like speeding up signals onboarding or simplifying account setup. We've also improved partner onboarding as well, which is helping us scale. And we've seen some improvements in the median login to spend time for that advertiser cohort. I also just want to recognize the business development team has been doing a great job onboarding more customers. So we'll definitely be investing there as well as we work to further accelerate the growth we're seeing with small and medium customers.

Thank you.

The next question comes from Michael Morris with the Guggenheim Securities. You may proceed. Thanks for taking the question. Good afternoon. Wanted to ask about direct response advertising. Can you share how much was the 8% growth in the quarter and acceleration from the core trend in the second quarter when we removed the impact of the execution error that you guys had? And then as you look forward, can you return to double-digit growth in direct response advertising? And if so, I know that you have a number of initiatives. I appreciate all the details, but would you maybe give us the top two or three contributors that can really impact that growth rate over the next year? And I've got to just slip one more in. Following the error that you did experience last quarter, can you just provide an update on your comfort and confidence with the stability of the bidding and optimization tools now to kind of ensure that you wouldn't have that happen again? Thank you. Derek Fox, LGO Admissions, hey there it's Derek speaking thanks for the question yeah direct response revenue was up 8% year over year in the most recent quarter, I was an acceleration of.

Derek Fox, LGO Admissions, Three percentage points over the prior quarter so we're pleased with the progress there. Derek Fox, LGO Admissions, What we saw is good strength in our pixel purchase demand, as well as the APP to optimizations and really broadly across the SMB segment helping to drive that acceleration in the quarter. You know, when you're looking at ad revenue broadly, you know, with direct response being the vast majority of it, we saw really good strength across Europe and rest of world. Europe in particular grew 12% year over year. That was an acceleration of six percentage points. Rest of world grew 13%, which is an acceleration of 10 percentage points in a quarter. So really strong results there, you know, both across the LCS and in particular the SMC market there. As we look at North America, that business still lagged a little bit and so dragged on the rate of acceleration on the overall business as well as in DR specifically. Within North America, though, really pleased with what we're seeing on the SMB segment, up to more than 25% year over year in Q3.

So given the strong momentum that we're seeing in Europe and rest of the world and with the SMB business globally, We're pretty pleased with what we're seeing on both the ad platform and our ad units there in terms of driving improvement on revenue and the business generally. I think if it pertains to our large client segment in North America, we saw a small decline there. We've been really focused on doubling down on what's working in the business, but also making targeted adjustments or go-to-market operations there in order to drive growth. We don't have know recovery in that north america large client segment really baked in you know in q4 obviously with a guy but we expect that the work that we're doing there will help us build momentum over time and if we can bring the growth in that portion of the business back up to to what we're seeing elsewhere then that is a path to further improvement in the overall growth and ads business going forward so hopefully that gives you you know a sense of what's driving the growth and acceleration on dr and Of course, we're, you know, watching our ad platform extremely carefully and the roadmap there and working with our teams to execute well there. And I think it's showing up in the results that we're seeing on the ad platform across the business globally. Thanks very much. Thank you. The following comes from Shweta Kajuria with Wolf Research. You may proceed. Okay. Thank you for taking my question. I just had a quick one on infrastructure costs for next year. Yes, could you please talk to your conviction level on keeping infrastructure costs basically flat next year and what, in your view, could drive those costs higher? And when would you think you would step in? Thank you. Sarah speaking here. You know, there's a number of different drivers here. Obviously, over the last several years, one of the really big drivers of our growth and infrastructure costs has been the rapid growth and investment of ML and AI infrastructure. And we do expect that we're going to be able to deploy capacity there, but we're getting a big focus on capacity utilization improvement. The other is we've scaled the business, obviously a lot with the growth in our community. And there's an opportunity for us to do work around the efficiency of that cost structure and our cost to serve. So both in terms of the services we're utilizing from our cloud partners, the pricing of those services, but then also just how we're engineering our product and the cost to serve, which Evan talked about a little bit earlier in terms of our ability to calibrate that cost to serve relative to each market and its long-term financial potential. And so we think across each of these vectors that there's a lot of opportunity for us to make progress on the infrastructure costs and make progress towards that specific goal we had stated of working to make infrastructure flat into 2026. Hopefully that gives you a little more color. Thank you.

The final question comes from Ross Sandler with Barclays. You may proceed. Great. Evan, just a question on spectacles. So there's been recent press reports about potential financial partners. And I think some of your peers have done partnerships with, you know, these manufacturing or distribution entities. So what's your latest thinking here? your AR software stack is fairly advanced versus the field for smart classes. So how are you thinking about leveraging software versus the hardware side? Just any updated thinking there would be great. Yeah, thanks so much for the question, Ross. We've got a really exciting year ahead here as we prepare for the public release specs.

And we've been thinking a lot about ways to accelerate our technical leadership in the space is a form factor obviously we've been focused on now for more than a decade and i think we've been able to really leverage our advantages in terms of lens core the huge ecosystem of lenses that have already been built the amazing developer tools in lens studio and obviously now the snap uh operating system that runs on the current developer uh version of specs so you know one of the things that we have been doing uh to you know create more optionality in terms of our ability to accelerate is that in putting a specs into their own standalone, a hundred percent own subsidiary, that'll give us some options as we think about, you know, potential partners to work with to, to accelerate our leadership here in the space as we prepare for the public rollout. So definitely some great opportunities to partner. We really believe that the killer use case for specs is lenses. And we've seen some incredible lenses that have been created so far. You know, almost weekly, it seems like developers are rolling out new and unique experiences.

So if you haven't gotten a chance to check out the latest, I'd highly recommend it. Thanks so much. This concludes our question and answer session, as well as SNAP Inc.' 's third quarter 2025 earnings conference call. Thank you for attending today's session. You may now disconnect.