AppLovin (APP.US) 3Q25 Performance Meeting: Accelerating the Introduction of AI Technology to Further Improve Advertising Efficiency and User Experience.

Recently, AppLovin (APP.US) held its 3Q25 earnings conference call. The company's core gaming business model upgrade in this quarter drove revenue to increase by 68% year-on-year to approximately 14.05 billion US dollars; adjusted EBITDA reached 11.58 billion US dollars, an increase of 79% year-on-year, with a profit margin of 82%, up 1 percentage point from the previous quarter, mainly benefiting from operational leverage and moderate relief from the impact of foreign exchange fluctuations. The cash flow conversion rate of adjusted EBITDA in this quarter reached 95%, slightly higher than the second quarter. Free cash flow reached 10.49 billion US dollars, an increase of 92% year-on-year. Due to not having to pay semi-annual debt interest this quarter (usually paid in the second and fourth quarters of each year), the free cash flow profit margin increased compared to the previous quarter. As of the end of the quarter, the company held cash and cash equivalents of 17 billion US dollars. The company expects total revenue for the fourth quarter of the 2025 fiscal year to be between 15.7 billion and 16 billion US dollars, an increase of 12%-14% over the previous quarter; adjusted EBITDA is expected to be between 12.9 billion and 13.2 billion US dollars, with a target EBITDA profit margin of approximately 82%-83%. Company executives stated that the company is accelerating the introduction of AI technology to further enhance advertising efficiency and user experience. Currently, the company is expanding its paid advertising capabilities to the Axon Ads platform and continuously optimizing the onboarding process and channel performance of new advertisers. In the future, the company also plans to introduce generative AI advertising creative to enhance user response rates and conversion efficiency. With a focus on long-term growth, the management will focus on optimizing the conversion process for new advertisers, expanding the promotion coverage of the Axon Ads platform, exploring automated customer acquisition models, and leveraging AI capabilities and increasing advertising density to enhance the overall growth momentum of the platform. Q&A Q: What are the main characteristics of the new advertisers signed since October 1? In comparison to the first 600 pilot partners screened last year, has their GMV decreased, does this imply a sinking user base? A: The new advertisers are primarily introduced through a recommendation mechanism, mainly in the retail format. Although the overall scale is slightly smaller than the pilot merchants rigorously screened last year, the gap is not significant, and the coverage categories are more diverse, with a roughly equal structure. Q: Regarding the 4Q performance guidance this year, how did you reference the seasonal data from the same period last year in e-commerce? What assumptions are made about the performance of existing customers and new e-commerce customers? A: Last year's fourth quarter was a period when e-commerce business was just starting and the growth rate was exceptionally high, so it is not suitable to make direct comparisons this year. The performance guidance is mainly based on the positive performance of the e-commerce referral program, the continuous optimization of the business model, the business updates mentioned in the third-quarter financial report, and the normal holiday seasonal factors. Therefore, the expectations for this quarter's performance are more optimistic than the previous one. Q: With the significant increase in conversion rates, does this mean that even with a significant increase in the number of e-commerce advertisers in 2026, there is no need to significantly increase the advertising impressions to support business growth? A: Yes. The company is currently providing a large number of advertising impressions to over 1 billion users every day. Under the net income model, increasing the conversion rate is key to business growth. The core paths include continuous model optimization, which has a huge impact on increasing conversion rates; and we are just starting to explore the application of neural networks and artificial intelligence technology to the advertising system. Although this field has not been around for long in creating real value for the industry, the potential is already very significant. In addition, expanding the density of advertisers can help the recommendation system tailor content more accurately; generative AI technology will be used in the future to automatically create creative materials adapted to the platform, in order to address the mismatch issue of some advertisers directly using short video materials from social media platforms, thereby further increasing conversion rates. Q: Beyond conversion rates, as e-commerce accelerates, do you expect growth on the supply side as well? Where will the main drives for supply come from - the demand for higher ad loads from developers, increased engagement in mobile games, or improvements in MAX platform functionality? A: This is the result of multiple factors working together. The MAX platform ecosystem is rapidly expanding, first, the improvement in advertising quality enhances user experience, preventing users from being disturbed by repeated ads while immersed in the games, and this experience optimization in turn enhances the density of demand for e-commerce, introducing more diverse content to both increase user retention and expand ad supply; secondly, the company is gradually unlocking developers who have mainly relied on in-app purchases for monetization in the past, or who have not advertised or have only advertised minimally, by providing them with more efficient monetization methods to expand the supply side resources; finally, with model optimization, the same developer can reach more high-retention users, achieve more efficient monetization, these factors together drive natural growth on the supply side. Q: The data of a 50% week-on-week spending growth for e-commerce customers is very impressive, can you provide some background information? For example, how was the performance during the pilot stage? How do you evaluate the current growth level? A: The business is still in its early stages with only one month since its inception. Customers need time to complete pixel integration and system deployment, usually requiring a period of over a week for onboarding. Despite the customer scale not reaching that of the pilot period last year, the growth rate is still very rapid, showing strong potential for compound growth. More importantly, this growth is not dependent on massive subsidies or sacrificing user experience. Customer feedback and content quality remain high, and we are focusing on optimizing the onboarding process for customers from registration to integration going live. The company hopes to expand the Axon platform to future customers and is supporting businesses to self-serve their advertising through embedded AI tools. This shows that our growth path is healthy and replicable. We are expected to officially open the platform in 2026 and truly attract a large number of advertisers in the following quarters and years. Q: When seeking the potential for multi-billion dollar business growth, how do you balance the demands of existing core gaming clients and new e-commerce advertisers? Will the introduction of new bidders affect the gaming advertising business? Can model optimization automatically solve these balance issues, or will you actively limit certain business growth? A: We will not intentionally limit growth, but let the platform evolve naturally. As the density of advertisers increases, it will not only squeeze out core gaming advertisers but also make gaming business advertising spending more efficient. The efficiency of improving conversion rates by solely relying on repetitive gaming ads in the past is very low and wasted many display opportunities; however, with the addition of e-commerce and other diverse advertisers, the model can more accurately match different user intentions, making better use of display resources. Although the number of ad impressions may decrease, the CPM of gaming clients will increase, resulting in equivalent or higher revenue. New advertisers bring more rich data, and the neural network model can further improve the accuracy of ad placements, benefiting both e-commerce advertising and helping gaming clients acquire higher-quality users. Ultimately, the platform will strengthen demand on both ends by expanding the profit pool while enhancing overall growth, rather than trade-offs. Q: About optimizing the advertiser onboarding process and generative AI creative tools, what is the current progress? How far away is the launch of one of these tools? Or how fast is the progress? Are these two tasks limiting factors for current launch or adoption, or are they crucial to expanding a broader customer base? A: The Axon Ads website has deployed keyword-driven chat Siasun Robot & Automation systems to optimize advertiser input and internal content review, and related functions are basically established and will be fully launched soon. We continue to integrate a variety of Siasun Robot & Automation functions into the system to achieve automatic quality checks and content standardization; as users and data continue to accumulate, the tools will continue to improve. In addition, generative AI ad creatives are currently in the testing phase, and with the rapid iterations of Sora 2 and Veo 3, we expect to enable AI-generated materials in a matter of weeks or months, significantly increasing creative output. However, these functions are not hindrances to the platform's openness; the real key is to ensure a smooth conversion funnel, a seamless user onboarding experience, and no operational impact from surging demand. Once these standards are met, the full platform openness will naturally follow. Q: With the transition from the App Store and Play Store's 30% commission to self-payment models, is it possible for revenue to materialize earlier than expected? Has this transition already affected the third-quarter performance? A: Currently, the financial contribution from the transition is not yet significant, and this impact will gradually become apparent over several quarters. While reducing the commission rate from 30% to single digits is unrealistic, even at a level of 15%, it will significantly increase the lifetime value (LTV) of some applications, allowing funds to be reinvested in content development, profit improvement, or marketing purposes. Although the external payment environment is determined by platforms and regulations and is beyond our control, we focus on optimizing our tools and models. The strong performance in the third quarter is driven by model iterations, template upgrades, and the synchronized growth of advertisers and ad placements, and we remain confident in the annual growth target of 20% to 30% for the core business, with the inclusion of the self-service platform providing additional growth stack effects. Q: Currently, advertisers are mainly concentrated in the mobile game industry. Will there be expansion to other media in the future? Considering the antitrust litigation faced by Alphabet Inc. Class C's advertising technology, if assets like AdX and Alphabet Inc. Class C ad managers are split up for sale in the future, will AppLovin consider acquiring such resources? A: Our core strategy is always to provide the best solutions for our partners. Currently, we have had excellent results in serving advertisers and game publishers. However, in the broader open network publisher and other app publisher fields, monetization efficiency is still low, and this is an area we need to further explore. On the other hand, CTV and other non-gaming scenarios face similar challenges in monetization, with demand still far from being met. Therefore, we see these areas as potential expansion directions. The key issue in the current industry is not a lack of supply but limited advertising demand; as long as we can effectively introduce more advertisers and maintain competitive platforms, we can benefit both advertisers and publishers by expanding the supply side while maintaining overall growth. Therefore, we plan to gradually expand into non-gaming media in the future, but the emphasis will remain on doing well in our existing business and making breakthroughs on the demand side. Q: With the growth rates of the in-app purchase market and in-app advertising market differing, have there been changes in how the Lifetime Value (LTV) of customers is calculated? Has the proportion of ad revenue in the overall LTV significantly increased compared to a year ago? What does this change mean for the company? A: The in-app purchase market is more mature, while the in-app advertising market is growing faster, and the growth rate of the MAX platform is multiple times that of the in-app purchase market. Established gaming publishers are leveraging more efficient monetization tools to drive growth, and new publishers are quickly adopting the advertising model, leading to continuous expansion on the supply side and increased efficiency in monetization. In comparison, the in-app purchase market lacks effective new monetization tools to drive growth still requires a reduction in the platform's cut to push for growth, but achieving this change uniformly between different platforms is challenging. Therefore, the company focuses on boosting promotion efficiency and revenue-generating capabilities for developers through the advertising model, especially within the MAX ecosystem, where supply growth is a key driver of overall market expansion and will synergize with other business sectors to accelerate overall growth. Q: Regarding the 50% week-on-week growth in e-commerce business, is this the key indicator for deciding the timing of a full public test? If not, what other conditions need to be met before full opening? A: While the 50% growth is exciting, it is not the sole indicator for determining a full opening. We are more concerned about optimizing the conversion funnel from registration to going live, especially when brand awareness is limited and the number of referral codes is restricted. That's why we are working to reduce the likelihood of user drop-off after registration. In terms of feature demands, there are currently not many feedbacks, perhaps because new customers are still adapting to the platform. However, there are relatively more feature demands from early customers, but more feedback may come with business growth in the future. Q: When considering paid marketing, you mentioned the possibility of increasing investment, yet based on the third-quarter sales and marketing budget, it seems there has been no significant advancement. How should we understand the timing points and scale of increased marketing investments in the future? A: We are currently conducting small-scale tests, and the key to increasing investment in the future depends on the performance of conversion rates, the comparison of user LTV and acquiring costs. Given the huge scale of our business, even with increased placement levels, the relative contribution to income is not high, and we have always adopted efficient performance marketing strategies, so there is no inefficient spending. Brand awareness is still low, therefore if tests show that conversion rates are healthy, we will accelerate placements to push users through the full conversion funnel automatically from registration to going live, helping maintain a lean sales team. Q: Will the conversion rate or revenue profit margin for e-commerce business differ from the core game business? Will early incentive mechanisms like advertising points cause short-term structural differences? A: There will not be significant differences. The advertising points we provide represent a very small proportion of the overall value of new customers, consider them as regular customer acquisition costs similar to what one would see in growth marketing, so the impact on conversion rates is very limited. Moreover, the platform operates under a unified auction mechanism, so when e-commerce or other category conversion rates improve, the entire platform benefits from the same commission rate. Therefore, a higher demand density is expected to lead to an increase in overall conversion rates and revenue. Q: As the company's core game business and e-commerce advertising business grow rapidly in sync, with the e-commerce sector's conversion rate possibly lower than the gaming business, do you fear that the advertising inventory in the core gaming market will gradually deplete due to the increase in conversion rates? Additionally, if there is a potential opportunity to acquire assets from Alphabet Inc. Class C advertisements, would the company consider participating? A: The company believes that the density of advertisers is still very low, and there is still a long way to go before the inventory is "depleted." Currently, the advertising spending scale for each quarter has exceeded 11 billion US dollars, but the corresponding number of advertisers is only a few thousand, while the platform's daily active users exceed 1 billion. Therefore, as the number of advertisers and ad density increases, data scale, model iterations, and conversion rates will continue to improve, and there is no need to worry about supply-side bottlenecks in the short term. As for supply expansion, the company is investigating monetization opportunities of broader publisher and connected TV markets, and at the right time, may simultaneously lay out supply and demand sides to strategically explore potential acquisitions of Alphabet Inc. Class C assets. However, the company is currently focused on building demand-side capabilities. Q: Did the third-quarter user performance become more active with tool enhancements? Additionally, could you provide some background on the PSU plan launched in October and its audience? A: Both customer advertising ROI and tool efficiency have been continuously improving. With the upgrade from the old dashboard to the Axon Ads self-service platform, user understanding and involvement have significantly increased, allowing various optimizations to take effect and enabling us to compete with long-established industry competitors, despite only launching the product for a year. The PSU plan launched in October is a talent incentive and reserves mechanism set up for the engineering team, aimed at retaining existing core talent and supporting future engineering talent recruitment and development. Q: Has all the referral codes been distributed already? Until when will the distribution scale continue? How do you define and filter out "low-quality" advertisers? A: The referral codes have not been exhausted yet, and we will distribute more codes dynamically based on the quality of customers recommended by partners to ensure high-quality referrals are not restricted, while also moderately controlling the pace to allow time for tool development. As for the advertiser screening criteria, we currently adhere to a high-quality entry mechanism, ensuring that the audience is seen as a stable large group and cultivating their high-quality awareness of our advertisements. In an environment lacking competition, where users are bombarded with singular advertisements when there is a lack of variety, we must ensure that every ad has high value to the customer. Q: From a macro perspective, with the core game business of your company growing rapidly alongside the e-commerce advertising business, and the possibility of lower conversion rates in the e-commerce sector compared to the game business, are you concerned that the advertising inventory in the core game market will gradually deplete due to higher conversion rates? Long-term, If Alphabet Inc. Class C ad assets present potential acquisition opportunities, would the company be interested in participating? A: The company believes the advertiser density is still very low, and there is still a long way to go before the inventory is depleted. Currently, the advertising spending scale has exceeded $11 billion each quarter, but the corresponding number of advertisers is only a few thousand, while the platform's daily active users exceed 1 billion people. Therefore, as the number of advertisers and density increases, data scale, model iterations, and conversion rates will continue to improve, and there is no need to worry about supply-side bottlenecks in the short term. In terms of supply expansion, the company has been exploring monetization opportunities in broader open network and net-connected TV marketplaces. When the time is right, the company may simultaneously lay out supply and demand sides to explore potential acquisitions, but the current focus remains on building demand-side capabilities.