Roblox Corporation (RBLX.US) recently held its fourth-quarter 2025 earnings conference call. The company stated it will continue its strong partnership with Tencent and recognizes significant opportunities in the Chinese market. Roblox has refined its perspective on the Chinese market: should it enter, the company would adopt a segmented deployment approach. The company's infrastructure is uniquely designed to be abstracted and deployed across multiple locations.
Regarding details on age verification and estimation, this primarily highlights the scale at which Roblox has become a cultural phenomenon. Although verified ages are slightly lower than user-reported ages, this actually confirms the platform's success. The company focuses on two main aspects: first, the user base covering those aged 18 and above is growing at a rate exceeding 50%; second, the technological advantages used to expand the under-18 user base are identical to those used for developing the 18-and-over demographic—from cloud services to applications, and further to content discovery and social graph vertical integration. The company firmly believes these technological strengths are equally applicable. More importantly, the technology continuously supports increasingly realistic user experiences. Consequently, the company remains highly confident in the growth prospects for the 18-and-older user group.
Q&A Session
Q1: Could you elaborate on the factors influencing the annual bookings revenue outlook? What is the visibility on the content schedule and releases? Also, what are your views on the opportunity with users aged 18+ and the latest progress in AI gaming?
A: We have numerous internal leading indicators that show some correlation with system health. This includes content diversity, content distribution, content velocity, and the types of content resonating across different age groups and genres. We see healthy trends in these areas. In Q4, we observed an increase in content diversity. As mentioned in the shareholder letter, experiences outside the top 10 grew faster than in Q3, both in terms of engagement and bookings revenue, and we've seen similar trends early in 2026. We appreciate the diversity and freshness of content. Even without blockbuster hits on the scale of "Grow a Garden" or "Steal a Brainrot," we see healthy growth from new titles, which gives us confidence in the platform's strength. I also referenced user quality in 2025. There was some uncertainty during that high-growth period: were these lower-quality users? But when we look at behavior on the platform, Q3 and Q4 new users act very similarly to our core users—in how they engage, spend, and retain. This gives us confidence for the business in the coming quarters. However, predicting the business position 12 months out still involves uncertainty. That's why we believe annual guidance isn't the optimal long-term approach. We used a slightly wider guidance range for 2026 to reflect this uncertainty. Importantly, our guidance does not assume another viral hit like "Grow a Garden" or "Steal a Brainrot." While we are optimistic about such occurrences, we do not incorporate them into guidance as they are unpredictable.
Regarding the AI future, we are optimistic about the expansion of how games are defined, and AI will drive this. Let me return to our mission: to connect a billion people daily with optimism and civility. I emphasize the word "connect," meaning bringing multiple people together to play, learn, and work. Since going public, we have shared our vision specification: many, many people gathering together to play in physically simulated environments. We began adding concepts of NPCs and people to make these environments increasingly realistic and places where people can create, modify, and change things in real-time. As we demonstrated recently on platform X, we are starting to use AI across the entire tech stack—using 3D upscaling techniques to make scenes more vivid and realistic, and training NPCs using 13 billion hours of video data and real 3D world data. We particularly highlighted breakthrough results from our internal world model team—they not only utilize video data but also integrate internal Roblox data to build world models. This innovative approach can be used for content creation and other scenario needs. I must emphasize a crucial point for our technical strategy. We are developing multiplayer platform technology focused on building social features that foster user interaction. Many current products operate in the video generation model space, whereas we focus on asynchronous 3D multiplayer cloud spaces. We advise investors to carefully understand this distinction; relevant posts can be referenced from our recent activity on platform X. Nonetheless, we remain optimistic about integrating various AI components to build our technology stack.
Q2: Regarding the evolution of the platform's discovery function over the past 12 months, what opportunities lie ahead based on this evolution? How does this align with the strategic priorities for the platform or individual products in the coming years?
A: The "discovery" function (game recommendation) is a difficult problem to get right, and we believe transparency is very important. We optimize discovery for long-term benefits, not short-term gains—this includes long-term user engagement benefits as well as long-term platform and creator health benefits. Any short-term optimization of "discovery" likely isn't optimal for long-term enterprise value and ecosystem health. Therefore, we consistently focus on how to personalize recommendations for each user in a way that connects great users with great content over the long term. The results we see are more good content reaching more great users, increased diversity in the content mix, and some future opportunities. Keep an eye on "Moments" (the discovery feature); we believe that for many users, browsing in-experience will increasingly complement our discovery function. This is a long-term journey, and we will continuously improve.
Q3: Quarterly gross margin was the second highest since 2020. Is this improvement related to direct payment transitions? Can you update on the progress of direct payment initiatives? Also, regarding potential threats from AI from other companies, what are Roblox's differentiating factors?
A: Regarding quarterly gross margin, I would point to two factors. First, as you hypothesized, we did get some tailwinds from COGS. As we've discussed before, we are experimenting within the product to guide users towards purchasing Robux on lower-cost platforms. This performed better than expected in Q4 and was very helpful for margins. Second, another source of quarterly margin expansion was bookings revenue growth. Bookings revenue was better than expected, providing strong leverage on fixed costs. Looking ahead, we continue to expect some improvement in the COGS rate as we can shift more business to lower-cost platforms. This may not be linear, but long-term, we see it as a source of business margin expansion.
Regarding AI threats, that's a good question. I'd like to shift to sharing how we internally view this—namely, the disruption opportunity in the opposite direction, which is the opportunity for the Roblox vision to expand from gaming to the future hybrid of entertainment. A decade ago, many in the market discussed how video could become interactive, but those experiments were difficult. Another way to think is how gaming can expand and become part of entertainment. We are confident in what we are building: multiplayer technology running in the cloud that can load increasingly faster and be consumed in smaller chunks. World models are interesting; we do think they have opportunities for walking around and painting worlds. There's a very interesting opportunity to think about where video ends, where snapshots end, and where interactivity begins. So we are developing our own models there, but the core of our technology will continue to be the very difficult problem of 3D cloud synchronization and building communication-like technologies.
Q4: How should we think about stock-based compensation expense for 2026? How does the flexibility of stock-based compensation help with long-term strategic investments?
A: Internally, we always run multi-year stock-based compensation models across a very wide range of stock prices and operate the business in a way we feel comfortable with. So we do consider these things and build multi-dimensional models. I would say we look at this from a fairly long-term perspective. Due to stock price movements, incentive expenses might peak at certain points. But ultimately, we believe we will create shareholder value, which will cause incentive expenses to decline over time. If you look at what happened over the past few years, this has absolutely been the case. So it's something we monitor, but we focus more on the operational results of the business, as dilution will be addressed through that.
Q5: What is the latest progress on the advertising business? How will advertising contribute to growth in 2026? And details on the rollout of age verification?
A: We are very excited and proud of how the age verification rollout has been conducted and are optimistic about its results. We believe it expands our team's thinking on how to long-term become a unique platform that can safely serve all age groups. We set an ambitious internal goal to ultimately roll out this feature frictionlessly. By doing so, we discovered many other optimization opportunities, and I am very satisfied and pleased with how the rollout proceeded.
Regarding advertising, we expect the advertising business to achieve healthy growth in 2026. However, it remains small. It is not a major contributor to revenue. As you've heard me say before, it will take some time to grow. I think the long-term opportunity is very exciting due to the platform's scale and engagement, but it requires time. We are executing cautiously, building the ad product, integrating the technology into the platform, working with creators to expand inventory, and we will continue to be very careful to ensure this business is built the right way.
Q6: There have been reports that Roblox might be interested in expanding its presence in China. Can you comment on this? Regarding user age, with more young users than previously reported, does this mean it's harder than expected for young users to age up?
A: Regarding China, we continue our strong partnership with Tencent and see significant opportunities in the Chinese market. We have refined our perspective on the Chinese market: if we enter the Chinese market, we will adopt a segmented deployment approach. We believe our infrastructure is uniquely designed to be abstracted and deployed across multiple locations.
Regarding details on age verification and estimation, a few points need clarification. First, this highlights the scale at which Roblox has become a cultural phenomenon. Although verified ages are slightly lower than user-reported ages, this actually confirms the platform's success. I mainly focus on two aspects: first, the user base covering those aged 18 and above is growing at a rate exceeding 50%; second, the technological advantages we use to expand the under-18 user base are identical to those used for developing the 18-and-over demographic—from cloud services to applications, to content discovery and social graph vertical integration. We firmly believe these technological strengths are equally applicable. More importantly, the technology continuously supports increasingly realistic user experiences. Therefore, I remain highly confident in the development prospects for the 18-and-older user group.
Q7: The penetration rate of the age check rollout seems to be progressing smoothly. But have you noticed any changes in behavior or engagement levels between users who have completed the age check and those who haven't? Could the derivative effects of the age check lead older users to perceive this as a quality-of-life update, thereby contributing to engagement levels?
A: Let me give an example of why we are excited about this. With the deep integration of the age verification feature with social interaction, we can more accurately match users of different age groups with suitable counterparts. What makes me most optimistic is that this age-stratified matching mechanism can not only accurately connect older users with younger users but also achieve efficient matching across different age groups. We believe this will become a potential driver for long-term growth. Precisely because of this, we refer to it as an industry benchmark. In fact, after we implemented this solution, another major gaming company announced it would adopt the same solution, and a social platform also announced it would implement it simultaneously. We firmly believe this will eventually become the mainstream model in the industry. As one of the first major platforms to adopt this solution, we are deeply proud.
Q8: The Q1 bookings revenue guidance indicates a sequential decline of over 20%, which seems to be one of the larger quarterly slowdowns. Is there a specific reason for this, or is it simply conservatism?
A: My view is quite different. This is a quarter where we currently do not have a major viral hit. Achieving 40% to 44% top-line growth without such a hit is something I think we should all be very, very excited about, as it indicates the platform's health. We said last quarter that growth would moderate as we began moving out of the period with huge viral hits. So I don't think this quarterly trend should be surprising. If anything, I think we should be very excited about the start of the year.
Comments