Dropbox, Inc. (DBX) Q2 2026 Earnings Report, Transcript and Summary

Thank you for standing by, and welcome to Dropbox's Second Quarter 2026 Earnings Conference Call. [Operator Instructions] I would now like to hand the call over to Sarah Schubach, Chief Accounting Officer and Head of Investor Relations. Please go ahead.

Good afternoon, and welcome to Dropbox's Second Quarter 2026 Earnings Call. As a reminder, we will discuss non-GAAP financial measures on this call. Definitions and reconciliations between our GAAP and non-GAAP results can be found in our earnings release and our earnings presentation posted on our IR website at investors.dropbox.com. We will also make forward-looking statements on this call, including statements about our future outlook for our third quarter and fiscal year 2026 as well as our expectations regarding our business, assets, strategies and the macroeconomic environment. Such statements are subject to known and unknown risks and uncertainties that could cause actual results to differ materially from those described. Many of those risks and uncertainties are described in our SEC filings, including our most recent report on Form 10-Q and forthcoming report on Form 10-Q. Forward-looking statements represent our beliefs and assumptions only as of the date such statements are made. We disclaim any obligation to update any forward-looking statements, except as required by law. I will now turn the call over to Dropbox's Co-Founder and Co-CEO, Drew Houston.

Thank you, Sarah, and good afternoon, everyone. Before I turn it over to Ashraf, I want to briefly address the CEO transition we announced in May. Ashraf and I are currently serving as co-CEOs. And after this transition period, I'll become Executive Chairman, and Ashraf will become sole CEO. We're taking a deliberate approach to the handoff, and I'll remain deeply engaged as Executive Chairman. I want to say a word about why I have so much confidence in Ashraf. When he took over our core business, there were real questions about whether we could change our trajectory. He made a series of difficult calls and the business has gotten stronger every quarter since. He's built a strong leadership team and brought a level of operating rigor that has made this company better. You'll hear the results directly from him in a moment. The next phase for Dropbox is about execution and turning the progress we've made into consistent, durable growth. Ashraf is the right leader for our next chapter, and I'm looking forward to supporting him and the team. With that, I'll turn it over to Ashraf.

Thanks, Drew, and good afternoon, everyone. Before I turn to the quarter, I want to start by thanking Drew. Dropbox exists because of his vision and leadership over the past 2 decades. He built one of the most recognized technology brands in the world, and I'm grateful not only for the opportunity to help lead this company, but for the trust he has placed in me to help write its next chapter. I joined Dropbox because I believed it has significantly more potential than the market and customers appreciated. Dropbox has over 18 million paying users, one of the most trusted consumer and business brands on the Internet, strong cash generation and a global infrastructure built over nearly 2 decades. At the same time, I saw a real opportunity to improve execution, modernize the product experience and return Core to durable, sustainable growth. Today, my conviction is even stronger. Over the last 18 months, we strengthened the leadership team, sharpened our execution and focused relentlessly on the fundamentals, improving conversion, onboarding, activation, retention, pricing and packaging and delivering a better customer experience. While we're still early in the journey, the results are encouraging. We've turned Core from a business that had been slowing for years into one that is once again demonstrating sustainable growth. There's still a great deal of work ahead and a few quarters don't define success, but we're putting points on the board and reinforcing the belief that has brought me here in the first place. The second thing that has strengthened my conviction is something I did not fully appreciate when I joined, and that's how valuable Dropbox's foundation would become in an AI-first world. Over the last 2 decades, we have built far more than a storage application. We have built one of the world's largest and most optimized content platforms, one responsible for storing, synchronizing, securing, searching, processing and governing hundreds of billions of pieces of content across multiple exabytes of data. As AI makes intelligence more abundant, trusted content becomes more valuable, not less. Every AI application ultimately needs content to reason over, permissions to respect, governance to reinforce, version history to rely on and infrastructure that scales globally and securely. Those capabilities we have been building for nearly 2 decades, and they're what allow us to turn AI into durable value for customers rather than a feature that is easy to copy. That realization has shaped how we think about Dropbox's future. Our priority remains exactly what it has been since I joined, continue strengthening and growing core and build on it. I think about that opportunity in 3 connected parts. First, we're bringing Dropbox services onto a common platform built around shared content, identity, permissions, search and AI. Smaller teams increasingly powered by AI can build richer workflows faster because they're building on capabilities that already exist instead of recreating them. That also means showing up where our customers already work. We've launched integrations with tools like Claude and ChatGPT. And even without much dedicated investment behind them, we've already seen over 150,000 users connect to the integrations, an early signal of how embedded Dropbox already is in the way people work. Second, we're embedding Dash intelligence directly into Dropbox itself. Rather than treating AI as a separate destination that customers need to learn or adopt independently, customers expect intelligence to be a native part of how they interact with their content, helping them find it faster, understand it more deeply, organize it more effectively and ultimately do more of the work around it, all grounded in the trust, permissions and context already built on Dropbox. One of our biggest learnings throughout building Dash is that customers respond most to AI that is grounded in their own context and helps them get their work done. That's informed how we think about Dash going forward. We've come to see the bigger opportunity as Dash and Core together, bringing that same in-context intelligence natively into Dropbox for all our customers, not a stand-alone product for a subset of them. The third is using those capabilities to build deeper workflows in the markets where we are best positioned to win. The goal is not to become a broad software suite. It is to go deeper in a focused set of areas where content sits at the center of customers' work and where our existing assets give us a genuine advantage. Replay, our video and media review and approval tool is a good example, and we're validating adjacent opportunities such as digital asset management and other AI-powered workflows that extend naturally from our platform. Let me make that a little bit more tangible with an example. Take Westchester Publishing, what started as a place to securely store and sync files has grown over time into the foundation for much of their business operations. The Core Dropbox platform they've relied on for years also powers a custom portal they use to collaborate with internal teams and external partners. On top of that foundation, they've adopted Dash to find and organize content, helping teams quickly synthesize information and draft materials while piloting agentic capabilities that automate previously manual and time-intensive workflows. We're also seeing that the infrastructure behind Dropbox has become increasingly relevant in an AI-first world. AI models need trusted content, source permissions, audit trails, governance, multiplayer functionality and workflow continuity to deliver real value. And those are capabilities we've been building for nearly 2 decades. As AI companies build new products, many want to leverage that existing foundation rather than recreate it themselves, and we're seeing this demand already with our ChatGPT and Claude integration. It's still early, but we believe Dropbox can play an increasingly important role as the layer that connects AI to trusted customer knowledge and infrastructure across the broader ecosystem. That, in short, is our strategy, continue executing to restore durable growth in core through foundational improvements that increase our baseline, use that stronger foundation to build increasingly intelligent workflows for our customers and grow the flywheel that made Dropbox successful in the first place and over time, put the platform we've built to work more broadly across the AI ecosystem wherever that creates real value. Turning to the quarter. Our Q2 performance was largely the product of the foundational work that we believe is so critical to returning to sustainable growth in the long term. The platform intelligence and workflow strategy I just walked through is what we believe compounds on top of that foundation and unlocks a higher level of sustainable growth over time. We continue to see positive year-over-year revenue growth in Q2, excluding FormSwift, and we added 96,000 paying users, our third consecutive quarter of paying user growth. We also exceeded our guidance on non-GAAP operating margin, achieving over 39% and generated $283.5 million of unlevered free cash flow. Within Teams, our continued investment in pricing, packaging, onboarding, checkout and activation translated into stronger conversion and Teams net new ARR grew sequentially. Within individuals, targeted retention initiatives along with Apple Pay, SEPA and a clear upgrade experience for customers approaching their storage limits all contributed to a stronger monetization. These are not isolated wins. They are the kind of steady execution that compounds over time and is returning core to sustainable growth. At the same time, we continue to build toward a smarter Dropbox with AI natively embedded in the experience. As the product has evolved, we are transitioning the rollout of what we previously called Dash in Dropbox to the next-generation smart FSS experience, which we are currently testing with a select group of customers. This evolution does not change our rollout time line, and we remain on track to significantly expand access to our base throughout the remainder of 2026. We will scale thoughtfully, validating customer value, engagement and business impact along the way. As we enter the second half of the year, our priorities remain clear. keep building on the momentum we have established in core with a stronger foundation to innovate faster, adding AI as a native in context capability across our product portfolio. That's the platform intelligence and workflows we believe will define Dropbox's next phase of growth, reaching more of the over 18 million paying users already on Dropbox and leveraging the same flywheel that made us successful in the first place. With that, I'll turn the call over to Ross.

Thank you, Ashraf. When I joined Dropbox, investors were asking whether our core business could grow again. Today, I think they're asking a different question, not whether we can grow, but whether that growth is durable and ultimately, how much we can sustainably grow over time. Q2 doesn't answer those questions completely, but it does provide another meaningful proof point. Ashraf laid out 3 connected parts to our platform and AI strategy. I think about how those translate financially in 3 phases. Phase 1 was simply returning our FSS product to growth. Over the past several years, we had increasingly shifted our attention away from our FSS product because we no longer believed it represented our greatest opportunity. What changed was refocusing on the fundamentals, things like pricing and packaging, onboarding, retention, checkout, and that work has returned us back to positive growth. Phase 2 is where I believe we are today, proving that growth is durable, not just a couple of quarter results. We're encouraged by what we're seeing, 3 consecutive quarters of paying user growth, Teams returning to positive license growth and improving retention, but we're not overstating where we are. What lies ahead is proving to you what we believe that we can build and enhance products that will provide value to our customers and drive growth higher. Phase 3 is where Ashraf's strategy to bring Dropbox onto a unified platform, embed Dash intelligence natively and build deeper workflows around our customers' content becomes increasingly important financially. As we do that, the question changes from whether Dropbox can sustain growth to how fast we can grow over time. One thing that has strengthened my conviction since joining Dropbox is recognizing that we spent nearly 2 decades building and running infrastructure and intelligent services such as our content processing platform that becomes more valuable, not less in an AI-first world. Agent or human, we believe there is no future where there's not a lot more content. As Ashraf described, Dropbox is far more than a storage application. It's a trusted content platform with capabilities around storage, synchronization, permissions, governance, search and content processing that become increasingly important as AI becomes embedded in how work gets done. We believe those assets give us a differentiated foundation to build on, both inside Dropbox and over time, potentially other companies can also build on our content platform. We'll pursue that opportunity the same way we've approached the turnaround of Core with disciplined execution and capital allocation. We won't scale investment because an opportunity is exciting. We'll scale it because customers demonstrate they value it and because it generates attractive long-term returns. Ultimately, our objective is to compound free cash flow per share over the long term through sustainable revenue growth and a strong margin profile, investing where we have the strongest right to win and returning capital to shareholders when that's the highest return use of capital. Q2 doesn't complete the journey, but it reinforces our conviction that we're on the right path. With that, let me turn to our financial results. Unless otherwise indicated, all income statement figures mentioned are non-GAAP and exclude stock-based compensation, amortization of purchased intangibles, certain acquisition-related expenses, workforce reduction expenses and net losses on real estate assets. Our non-GAAP net income also includes the income tax effect of the aforementioned adjustments. In Q2, revenue increased 0.9% year-over-year to $631.5 million. Excluding FormSwift, revenue grew 1.7% year-over-year. On a constant currency basis, revenue excluding FormSwift, increased 0.1% year-over-year. Relative to our guidance, the outperformance was driven primarily by improving core FSS trends. Total ARR was $2.566 billion, up 1% year-over-year. Excluding FormSwift, ARR grew 1.7% year-over-year or 0.2% on a constant currency basis. We exited the quarter with 18.19 million paying users, a sequential increase of approximately 96,000 ahead of our expectations coming into the quarter. The outperformance was largely driven by outperformance in our single SKU. We also saw positive Teams license growth as a result of our ongoing pricing and packaging initiatives. Average revenue per paying user was $139.68 compared to $138.32 in the year ago quarter, driven by FX rate tailwinds and shift to more monthly plans. Gross margin was 81.6%, down roughly 60 basis points from the year ago period, primarily as a result of compute costs associated with rolling out additional AI capabilities to our teams base. Operating margin was 39.7%, ahead of our guidance of 38.5% and down roughly 180 basis points from the year ago period, driven by the gross margin dynamics I just described as well as increased marketing investment within our core business, reflecting a return to more normalized spend following the targeted reductions in performance marketing we made in the year ago period. Relative to our guidance, the outperformance was primarily driven by higher revenue as well as some timing-related savings shifted to the second half of the year for brand spend and outside services. Net income was $170 million compared to $197.7 million in the year ago quarter, with the decrease primarily due to higher interest expense related to our term loan facility. Diluted EPS was $0.75 compared to $0.71 in the year ago quarter based on the 226.8 million diluted weighted average shares outstanding compared to 276.7 million shares in the year ago period. Cash flow from operations was $238.5 million compared to $260.5 million in the year ago period. The year-over-year decline primarily reflects an increase of $30 million of interest payments, net of the associated tax benefit related to borrowings under our term loan facility. Capital expenditures were $3 million. Unlevered free cash flow was $283.5 million compared to $276.4 million in the year ago period. Unlevered free cash flow per share was $1.25 per share, up 25% year-over-year. Turning to the balance sheet. We ended the quarter with cash and short-term investments of $1.114 billion. During the quarter, we completed a new $400 million revolving credit facility, further strengthening our liquidity profile. The facility remains undrawn at quarter end and provides additional balance sheet flexibility. We also announced a new $900 million share repurchase authorization, reflecting our confidence in the business and reinforcing our commitment to long-term shareholder value creation. In the second quarter, we repurchased approximately 12.6 million shares, spending approximately $315 million. As of the end of the second quarter, we had approximately $1.385 billion remaining under our existing share repurchase authorization. I'll now offer our outlook for Q3 and our updated outlook for the full year 2026. For the third quarter of 2026, we expect total revenue to be in the range of $627 million to $630 million. Excluding FormSwift, this implies roughly flat year-over-year growth at the midpoint. We are expecting a currency tailwind of approximately $6 million. On a constant currency revenue basis, we expect total revenue to be in the range of $621 million to $624 million. We expect our non-GAAP operating margin to be approximately 38.5%, and we expect diluted weighted average shares outstanding to be in the range of 223 million to 228 million shares. For the full year 2026, we expect total revenue to be in the range of $2.513 billion to $2.523 billion, an increase of $13.5 million at the midpoint of guidance. Excluding FormSwift, this implies 80 basis points of year-over-year growth at the midpoint. We are expecting a currency tailwind of approximately $31 million. On a constant currency revenue basis, we expect total revenue to be in the range of $2.482 billion to $2.492 billion. We expect gross margin to be approximately 81.5%. We are raising our non-GAAP operating margin guidance by 50 basis points to be in the range of 40.0% to 40.5%. This implies an increase of approximately $18 million at the midpoint of guidance. We are also raising our unlevered free cash flow guidance, which we now expect to be at or above $1.070 billion, an increase of $15 million. We continue to expect CapEx to be in the range of $20 million to $25 million in addition to finance lease lines to be approximately 4% of revenue. Finally, we expect diluted weighted average shares outstanding to be in the range of 226 million to 231 million shares. I will now provide supplemental information as it relates to guidance. In Q2, we were pleased with our performance on paying user growth and continue to expect positive paying user growth for 2026. For ARPU, we expect modest sequential declines throughout the rest of the year. As I mentioned last quarter, our gross margin guidance assumes modest pressure this year from embedding Dash intelligence natively into Dropbox and expanding across our Teams base, partially offset by infrastructure efficiencies. Going forward, our gross margin profile will continue to depend on rollout pace, customer adoption and optimization work. So we continue to expect some quarter-to-quarter variability. We're increasing our operating margin and unlevered free cash flow guidance relative to our prior guidance as a result of Q2 performance and expected performance in the remainder of the year. As we touched on last quarter, we will continue to realize efficiencies within our R&D organization as we bring Dash and Dropbox closer together, giving our teams a shared foundation so they can build and ship faster with AI. Additionally, we see an opportunity to evolve and improve our go-to-market team and execution, and we are in the process of rebalancing that organization to focus resources on our priority markets, segments and routes to market, which we believe will drive greater efficiency and productivity through the remainder of 2026 and going forward. Lastly, we expect our full year weighted average shares outstanding to increase to approximately 226 to 231 million shares as a result of an increase in our 30-day trailing average share price. With that, operator, please open the line for questions.

[Operator Instructions] Our first question comes from the line of Rishi Jaluria of RBC.

Nice to see some kind of sustainable potential return to growth here. And Ashraf, welcome. Looking forward to working with you. Maybe two questions from me. First, as you think about the kind of cadence of paying users being added, third consecutive quarter, and kind of expect that to continue for the full year. Can you help me understand, with kind of declining ARPU, when can that start to show up in kind of the top line and drive maybe further acceleration from here? And then I've got a quick follow-up.

I think -- Rishi, it's Ross. So I think, number one, the ARPU commentary was modest decline throughout the year. So there's two opposing forces, FX and a little bit more mix of monthly mix to go up, and then the rolling off of FormSwift, which will end this year, and incremental Simple plan users make it come down. So when we kind of weigh that in the balance, we expect a modest decline in ARPU, not a major one. And on the users, I think, again, we -- as you pointed out, we've seen several quarters now of improvement there. We're seeing that on the individual side. We're also now seeing that on the team side for the first time since, I think, 2024, that flipped positive. So we called out that we expect to be positive for the year. We're being -- I think there's -- I'd just say there's a lot to do. Our initiatives are working. We think we can keep driving that forward, but there's a lot more data points and signal we want to see before we give more specific guidance around the growth there. So I think all in all, I think we're going the right way with respect to continued growth in net new paying users, which will drive ARR growth.

Okay. Got it. Helpful. And then if I think about some of the stats you've shared on Claude and GPT integrations, can you help us understand how is that translating into the business, whether that's user adds, whether that's greater stickiness? And then it feels like there's an underappreciated opportunity in that partnership. I think any of us that's built on Claude Code or Codex recognizes the value of having that connectivity to kind of a source of truth in all your content. So can you talk about what that kind of partnership and relationship and integration, how that could evolve over time and maybe become even more incremental to the business?

Thank you for the question. This is Ashraf. First of all, I think, if anything, this reinforces how -- what we talked about in the call, that AI is adding the need for storage and the ability to organize your files and find them and have a structured way for you to work. And so that's what we're seeing here. These are customers that are organically discovering Dropbox using the app, they're mainly using to find content, repurpose it and then later on storing it back on Dropbox. Our perspective is we want to meet customers where they are. We have a lot of amazing capabilities we're launching for our customers inside Dropbox, but we also want to meet them where they are. So if they want to work in ChatGPT or in Claude, we want to be able to offer them something there. And I think that's a natural extension of what we do. So think about an example where you're bringing content, trying to repurpose it, then you want to store it back. And at some point, you want to send it to someone and collaborate with the person you're sending it to. This is where we see Dropbox comes fully back into play because we offer deeper workflows at that point. And so what we're seeing actually is very encouraging, not only that this is growing organically, but also the engagement level and retention numbers that we're seeing are pretty encouraging. And if anything, it's validation that Dropbox has a much bigger role to play in this AI world than people appreciate.

Our next question comes from the line of Steve Enders of Citi.

Ashraf, good to hear from you on the call. Maybe just to start, maybe digging a little bit into the product strategy and what that looks like moving forward. I guess it would be great to kind of understand a little bit more kind of your view on what the future of the Dropbox product looks like, how you think about expanding the TAM into some kind of more specific areas. And I think you made a comment about wanting to create a platform that others can build on in the future. And so I would love to kind of understand what that looks like and what that entails.

Yes. Of course. I mentioned this is something that I didn't fully appreciate until recently. And this is something that we saw as we built our own agentic capabilities inside Dropbox. So I'm actually going to ground it maybe with an example and then give you how that looks differently than anything you've probably seen in other places. So today, this is like a real live example. You could -- imagine you're a project manager trying to get a marketing campaign off and running. You're going to need to find all the files. So our capabilities now enable you to find them semantically. You can even drop in a screenshot from a peer that sends to you something and say, find me that file, and it will find it. You put in a folder, it's still disorganized. You can then ask our agentic capabilities to auto-organize it. And then you find that the images, just like most of our customers have, it's called Image 3421. You can say, name it appropriately, and it will understand context and name it runner on a track or red car in a showroom. And so at that point, you're doing all this work and you want to actually start to loop in others and you want to tell them what you've done. You can actually -- because we have audit trail, you can actually summarize the changes and send it to your peer, your manager and say, here's the structure that I've created. And not only that, because we have connectors, you can just send that e-mail as well. Anything that happens in the product, because we have a file system, you can undo. So people can work safely. They can grant access to the agentic workflows to specific folders. So we have a tremendous history of building something that has permissioning, version control, audit trails and the ability to share securely, and all these things are even more needed in a world where agents are working. So I think of it as we've talked a lot about durability of the core business. We have 18 million paid subscribers, a massive distribution network beyond amazing what Drew has built with this company. And so I see it as we're going to bring these capabilities to provide deeper workflows in a way that saves customers a lot of time. This example that I mentioned takes hours and hours. You could do this in under 10 minutes. And so we believe that unlocks value. And this is, by the way, one example out of many. So you'll see us focus on engaged, the most engaged customers in marketing, creative, architecture, engineering, construction that rely on Dropbox for their work and content. And you're going to see us add very focused capabilities that make their lives a lot better. And you're going to see us add deeper workflows, and I think that's a new frontier for growth for us. And so in addition to this, as we started launching these capabilities, we realized that there might be an opportunity here for us to lean in on enabling other companies to leverage all these capabilities, and this is something we're excited to validate over time, but I think it's also very, very promising.

Okay. No, that makes sense, and that's great context. Maybe on just the constant currency revenue raise. I guess I would like to get a little bit more, I guess, like specificity and like what are the areas that maybe got better this quarter that you have line of sight to, that you're flowing through into the rest of the year? And I guess, how should we think about maybe the puts and takes on some of those components?

Steven, it's Ross. Thanks for the question. Yes, I think pleased now three quarters in a row, we've been able to beat and raise our revenue guidance. The growth rates are going up. I think all of us, yourself included, we want to keep seeing them go up. And internally, we're very focused on that. So everything that we've been talking about started when I got here in December with work we were doing around individuals very comprehensively around how we attract new users, how we convert them, how we retain them better. I think that started to pay off first. And then we talked about moving to teams as we entered this year and doing similar work around teams, and now you're seeing that reflected in the positive paying users. I'd say just on that front of the optimization across the customer life cycle, there's things we've already put in market that we still have visibility to paying off, and then there's new things to come. So that's not sort of exhausted itself. And so we are seeing improvements across both individuals and teams, again, across the life cycle, top of funnel conversion and retention. So that's reflected. And then not yet reflected is some of the things that Ashraf was talking about, which is how do we ultimately -- ultimately getting to that higher level of sustainable growth is about how we build the products, how we weave in the AI, all the Dash intelligence capabilities and just provide a lot more value for our customers. The cool thing is the AI example that Ashraf provided is already in Dropbox. Like we're already seeing it, we're using it and we're starting to roll that out. We're going to roll out these capabilities to the majority of our teams base for the rest of this year. So we're going to start to get more usage and then post that, hopefully, monetization. So that's not baked in yet, but those are more legs of growth to come. So it's early. We've got a lot that is working, but we also have a lot of work ahead and a lot to do. So we're trying to take a measured approach to how we think about guidance.

Our next question comes from the line of Matt Bullock of Bank of America.

Awesome. Welcome, Ashraf. I appreciate the color you guys provided on Phases 1, 2 and 3 of the Dropbox transformation. I guess I was hoping you could elaborate on how you think about the timing of moving between Phase 2, where we're at today, to Phase 3. And assuming you're able to execute against that strategy, how should we think about how that plays out across paying user and ARPU growth? And I guess maybe just a follow-up to that would be, what do you think the largest gaps are? Is it in product or go-to-market to execute against that transition?

Thank you for the question. So the phases Ross outlined was, one, to prove that we can get to growth. Second one was to make that durable. And the third one is to take that to the next level and expand significantly the growth rates that we have. And he highlighted that we're in the middle of Phase 2. I think that we have a lot of signal that what we're doing is durable, and we're excited about that. And there's a lot more that we're still doing around teams formation and expansion, and that becomes foundational. I think the biggest opportunities to get to Phase 3 is this deeper value for customers. I think that the thing we have to demonstrate that we can create significantly more value within our product for our most engaged customer base. And that translates into higher willingness to pay for additional SKUs and add-ons. So something like an attach rate, you would see it in things like that. So we do expect to see it in ARPU. I expect to see it in customers converting to paid more and then the ARPU going up as well. And that's the thing we're solving for by making Dropbox much smarter with the AI capabilities that we're rolling out that are focused on our most engaged and largest customer base today. So these are -- that's the Phase 3. And I think we're going to be testing that out this year. We're moving fast to make that a reality.

And Matt, just to add, it's Ross. We're not baking that in really for this year. So -- because you asked about like timing. So it's -- we're not going to comment on like the precise timing of like Phase 2 or 3 conceptually. But I think as you're hearing from Ashraf, like we're already putting the AI into the product. We're rolling it out this year. The application-level product builds that we're already underway on that. So the stuff is happening. We're not taking risk from it this year. And as we move forward quarter-to-quarter, we'll talk more about how that's progressing.

Really helpful. And then just one more, if I could. Dropbox Simple, it's been a really nice source of upside in the past couple of quarters. Can you maybe just give us an update on the scale of that SKU in the installed base today? And what's been so effective about it, whether it's better top of funnel, preventing outright churn events from higher-priced SKU users? Just trying to understand what the source of strength has been there.

I can start with part of the question, and Ross can chime in. This is Ashraf. So Simple was created on our end to make sure that we meet customers where they are and offer them the value that they expect. So what we found with individuals specifically in mobile -- actually on our mobile app is that they wanted lower tier storage, something -- that it's a little bit more affordable. They can just plug in their files and use that on the go. And so we met their demand by launching the Simple product, really, by listening to customer demand. And that's paid off, I think, always listening to customers and staying close to their needs. This is a good example where we launched it and it took off. And we're excited about the potential of that continuing to grow and how it even can boost the business and having that be part of a top of the funnel as well. As to specific performance data, I'll let Ross chime in.

When we think about the net new paying user commentary, you remember, it's a net number. So it's getting the benefit of both our improvements around top of funnel as well as retention. Simple is the largest contributor to the growth in net new paying users, but also as we talked about, teams is now positive and is a contributor. But the cool thing about net new paying users is it's broad-based. So it's individuals now, it's teams. It's other products that we have on the platform are also contributing, and it's top of funnel conversion as well as retention. So it's not like we're just relying on one thing to drive that. We've got a more broad-based set of inputs to help push that number.

[Operator Instructions] Our next question comes from the line of Jaiden Patel of JPMorgan.

Building on the last set of remarks, with this 96,000 paying users being what looks like the best in about 3 years, can you walk us through what the upgrade path looks like for these new paying users?

So let's start with the 96,000 users and how that's something we're really excited about. This came from very methodical changes we made. So to give you an example, we shrank the onboarding steps from 12 to 4 when we started a team. We enabled a lot of onboarding activation. We changed the product to recommend the next best actions. So all these things were instrumental to not just get top of funnel in absolute numbers go up, but actually the base is already coming organically, how we activate that a little bit better with the existing product capabilities that we have. So this was a big part of what we've done to get to that number. Now as you look at what we're doing over the next several quarters, we're investing deeply in multiproduct discovery, meaning you are able to see additional products and capabilities and features as you land on the main product experience. And we want to see that attach rates go up. So being able to sell you more, being able to, in context, say, by the way, we have a video review capability, we have a video. So you'll see us cross-sell and upsell a lot more. In terms of basic upsell path, we have a lot of optimization that we've done. So for example, when you get near your quota in storage, we're doing a lot more work than we've done in the past to tell you, by the way, you're nearing quota, here's options for you. We're telling people who are individuals that exhibit a team-like behavior. By the way, we have a team plan for you. It seems like you're inviting people in a business domain. So these are specific examples of how we're actually upselling and cross-selling in-moment. And I think that's something that I'm continuously excited about because I think this is -- when you talk about increasing ARPU, these are actual examples of where you're going to see this.

Great. And then another one on -- you mentioned gross margin pressure due to compute costs associated with rolling out AI capabilities. Is there a way to think about the floor for gross margins as you continue to roll out these features?

Jaiden, thanks. It's Ross. So for everybody, just keep in mind, there's two opposing forces on gross margin. One is the rollout of the AI functionality. We've been rolling out the first half of the year, but that will increase in the back half, and we expect to roll out the majority of teams in the back half. So it's not a perfect science to gauge timing of rollout and like adoption and usage and all that. So we're making our best estimates for that. But that would obviously increase cost of goods sold and weigh down margin. But what you guys also need to know is that there's an opposing force that goes positive, which is efficiency gains. And there's not just one, there's multiple. We have a really great infrastructure team that, number one, has great relations with the supply chain, is in front of the purchases and some of the pricing, but also just how we're running and optimizing that infrastructure and the systems we're running on. So they continue to deliver results that are improving efficiency that are counterbalance to the AI rollout. And remember, in the future, like we hope to monetize more things, including the AI product, which would be revenue on top of those costs. So I think like kind of implied in your question, a floor is like how low could it go because it's been going down. Right now, I would focus on the guidance for the year of 81.5%. As we get into next year, we'll talk more. But I just don't want people to assume that it has to keep going down because of AI costs when we do have other levers that we're pulling to help offset those costs.

Thank you. I would now like to turn the conference back to Sarah Schubach for closing remarks. Madam?

Thanks, everyone, for joining us today. We're looking forward to speaking with you next quarter.

This concludes today's conference call. Thank you for participating. You may now disconnect.