Applovin (Minutes): The self-service advertising platform will be launched in October this year.

The following are the Minutes of the FY25Q2 earnings call for$AppLovin(APP.US) , organized by Dolphin Research. For earnings interpretation, please refer to 《Applovin: What will the advertising kingpin use to amaze the audience?》

I. Review of Core Financial Information

1. Business Adjustment: At the end of this quarter, the Apps business was sold to Tripledot Studios. Relevant financial data is included in discontinued operations, with explanations and outlooks provided solely for the advertising business.

2. Equity Related: This quarter, 900,000 shares were repurchased and canceled, costing $341 million, covered by free cash flow, reducing the diluted shares outstanding from 346 million last quarter to 342 million.

3. Cash Flow:

a. Adequate Cash Reserves: At the end of the reporting period, cash and equivalents amounted to $1.2 billion, including $425 million net cash from the sale of the business.

b. High Growth in Free Cash Flow: This quarter's free cash flow was $768 million, up 72% year-on-year; slightly lower than last quarter, mainly due to timing differences in semi-annual bond interest payments and related taxes.

3. Q3 Outlook: Guidance for the third quarter of 2025: Advertising business revenue is expected to be $1.32 billion-$1.34 billion, adjusted EBITDA is expected to be $1.09 billion-$1.1 billion, with a profit margin target of 81%.

II. Detailed Content of the Earnings Call

2.1 Key Information from Executive Statements

1. Overall Performance:

a. The performance in the second quarter of 2025 was excellent, mainly driven by the continuous growth of the gaming advertising business. Growth sources include technological improvements, increased market demand, and supply-side expansion.

b. The MAX market continues to provide supply for the company and the entire market, serving as the core engine of growth. Although specific growth rates were not disclosed, the MAX market has consistently achieved double-digit growth, far exceeding the in-game purchase market.

2. Proprietary Model: Continuous optimization supports sustainable growth beyond industry averages. The company continues to expand its leading position in the industry, confident in achieving 20%-30% growth annually in the core gaming market.

3. New Opportunities: The current highlight is expanding new opportunities beyond the core gaming market. AXON Ads Manager (self-service advertising platform) has been quietly launched, laying the foundation for platform upgrades and long-term growth. The platform has the following advantages:

a. Promotes self-service operation and real-time management by advertisers, reducing operational barriers and process friction.

b. Supports credit card payments, optimizing financial processes.

c. Paves the way for automated, intelligent advertising operations (future capabilities include automatic ad generation and automated workflows).

d. Provides more convenient Shopify integration, simplifying customer onboarding.

e. Deep integration with attribution service providers, making reporting and attribution more accurate.

Starting October 1, 2025, the AXON platform is expected to open through a referral system, positively driving holiday season growth.We expect this to rapidly increase the number of advertisers, while allowing us to observe how advertisers enter and scale on our products through self-service.If all goes well, we plan to fully open the platform globally in the first half of next year.We believe that as the number of advertisers grows, especially in categories outside of gaming, you will see significant improvement in the numbers we can report.

4. Strategic Determination:

a. Over the past decade, we have continuously deepened our efforts in user scale, optimization technology, and platform capabilities, now capable of serving over 100 million users globally.

b. AXON will be launched as a new independent brand, and paid marketing will be initiated once the platform is fully open, forming sustained compound growth.

c. The company has a deep accumulation in efficient, performance-focused advertising products, with a highly diversified customer structure and significant future growth potential.

2.2 Q&A

Q: You decided to start paid market promotion next year to acquire new advertisers. What is the reason for this decision? Previously, growth was mainly achieved through word-of-mouth and industry self-propagation. Why is now the right time to start paid marketing? Through which channels will you mainly conduct market promotion in the future? How do you evaluate the return on investment of these market inputs?

A: Currently, our advertising solutions, whether serving e-commerce advertisers or broader industry categories, perform very strongly. But in fact, our penetration rate among global advertisers is still very low.

Our core idea is: If the model can perform so well with such low penetration, imagine the potential when we can truly open services to all SMEs globally. Our goal is to help businesses of any size efficiently acquire customers. If we can achieve this, we can achieve our goals.

The reason for doing performance-based market promotion is because our business model is extremely competitive—you can see this from our cash flow performance. Additionally, we are very good at performance marketing and have the ability to recruit advertisers using our own platform traffic.

You can imagine that in the future we will advertise on Facebook, LinkedIn, TikTok, guiding users who are playing games and have SME needs to our platform. Our model aims to remain lean and highly automated. If we can achieve full automation from advertisers seeing ads, registering, going live, to scaling, our LTV/CAC will be very healthy, without the need to massively expand the sales team.

Q: You mentioned that the supply side of MAX remains strong despite the current high base. Is this growth mainly from further increasing share in the mediation market or from increased ad impressions from existing customers? What are the main sources of supply-side growth?

A: We often talk about the mobile gaming market, many people focus on the in-app purchase market, but our growth actually comes from technological advancements and increased demand, which not only pushes up CPM but also expands supply. The MAX mediation platform already has high penetration in the industry, so there is limited space to further capture other platform shares. We want to emphasize that even so, the overall audience and usage of the platform continue to maintain double-digit high-speed growth, far exceeding the usual 3%-5% growth rate of the gaming market.

Secondly, the growth of the MAX market is not simply relying on "share increase." As advertising platforms advance, the overall market size is also growing. The number of users playing games every day is constantly increasing, and ad inventory grows accordingly, benefiting us and other participants on the platform. For us, the vast majority of transactions can charge MAX platform fees, and our own DSP is very profitable when winning inventory.

Q: AXON is about to launch, how do you view its potential impact on the overall business? Will it become an important factor driving performance? Can you describe the scale and prospects of this opportunity?

A: In the fourth quarter of last year, during the e-commerce category pilot, we quickly expanded and attracted hundreds of advertisers, driving rapid growth in Q4 and Q1. These new types of customers contributed significantly to business increments. Subsequently, we also clearly needed to enhance platform functions, such as building ad management tools (Ads Manager), supporting dynamic product ads, further integrating with attribution companies, and launching Shopify integration applications. To ensure product quality, we proactively slowed down the pace of advertiser onboarding in the past few quarters.

Now, Q4 will first open the self-service platform through invitation referrals, meaning existing customers can invite peers to join and achieve self-service advertising. We expect this to quickly increase the number of advertisers while allowing us to see the full chain performance of the self-service process. If all goes well, it will be fully open to a broader market in the first half of next year.

Especially in categories outside of gaming, as the number of advertisers grows, the business increment space will be very considerable.

Q: Is your statement about the e-commerce business accounting for about 10% still accurate this year? Has the recent restriction on e-commerce customer onboarding affected overall growth this quarter? Would growth have been faster without the restriction on e-commerce advertisers?

A: We have indeed suppressed business growth by limiting the number of advertisers, currently focusing on enhancing the existing advertiser group. As Matt mentioned earlier, most of this quarter's growth still comes from gaming advertisers, with quarter-on-quarter growth of about 9%, the gaming industry still maintains a high growth rate of 30%-40%, far exceeding our long-term target of 20%-30%.

In terms of e-commerce, after experiencing rapid growth in Q1, we proactively slowed down new advertiser onboarding due to the need for function enhancement. Currently, the e-commerce business accounts for roughly the 10% range we set at the beginning of the year, and due to the large increase in gaming business, this proportion may not have risen significantly. In the future, with the arrival of the Q4 holiday season, existing customer advertising will increase significantly, while the speed of new advertiser onboarding will create a historical high. It is expected that the e-commerce business will achieve significant improvement during the "soft launch" period, and subsequent global formal opening will further accelerate growth.

Additionally, we currently also limit overseas audience purchases, with the vast majority of users actually outside the United States. After fully releasing international market traffic on October 1, the overall e-commerce business will usher in a new round of growth.

Q: Besides Unity, can other companies also access and use game engine data? Do your customers have their own game engine data, and is there a possibility for you to access and utilize this data in the future? Do you think game engine data is important for advertising and platforms?

A: When we integrate into applications as a platform, both publishers and advertisers can access a large amount of user behavior data, and our models (such as Axon 2.0) perform exceptionally well in this regard. Our market penetration rate in the gaming industry has already exceeded 70%, providing great insight. The key to future business is whether we can extend these capabilities to all industries and various customers. If achieved, we can more comprehensively understand consumers, whether they are gaming users or not.

It is important to emphasize that we cover over 1 billion users, and many are not just playing games. Compared to the entire consumption scenario, gaming is only part of their spending. As we enhance our understanding of broader consumer behavior data, advertising prediction and effectiveness can also be greatly improved.

Q: When existing advertisers participate in the referral program, will there be referral rewards? Are there restrictions on referral targets? After promoting to international inventory, can international advertisers also participate in the referral program? For those advertisers who have been waiting in line for a long time, if there is no referral, is there a priority opportunity to participate?

A: The referral program will be continuously adjusted and optimized at the initial stage. Overall, we do not plan to provide material rewards for referrals. Our customers themselves benefit greatly from the platform, users enjoy using it, and there are already many spontaneous positive shares. We expect that when customers have invitation permissions, they will proactively recommend peers because the platform itself has always been relatively scarce and has barriers, and this kind of spontaneous dissemination will be healthier.

As for those advertisers who have been waiting in line, they will most likely still need existing customer referrals to be automatically approved to enter the platform.

Q: Your guidance for the third quarter appears to be slightly accelerated compared to previous quarters. Does this include the impact of the Apps business sale? Can you quantify the contribution of the Apps divestiture to performance growth?

A: One difference we made this quarter is that we included the revenue from the Apps business divestiture in the Q3 guidance, which will bring a slight revenue increase and has been included in the Q3 performance outlook.

Q: You are developing two AXON models, one for gaming and one for e-commerce and broader industries, right? How long do you estimate it will take for the newly developed model (such as the e-commerce model) to achieve the same effect as the existing gaming model, and how much data accumulation is needed?

A: Currently, it is impossible to determine how long it will take for these two models to achieve the same effect, as it depends on future data accumulation.

The architecture of the two models is very different, for example, user behavior on websites and app stores is completely different, and we cannot directly measure the App store. For example, when Axon 2.0 was launched, we already had a penetration rate of 50%-60%, and now it has almost become an industry standard, with higher penetration. In contrast, e-commerce is just starting, although hundreds of e-commerce companies have already launched, but compared to the overall industry scale, it is still small, and there is huge space for data accumulation.

However, the data collected by the platform is not limited to a specific industry, and cross-industry data enhancement will also drive the overall capability of the model. For example, users who can afford high-end products are likely "big R" in games, and this cross-industry insight will continue to increase. Overall, as the platform's openness increases, data and demand will form a positive cycle, driving model progress and business growth every quarter.

Q: For the targeting and optimization needs proposed by Web advertisers (such as excluding existing users), are these functions you have already encountered and solved on the gaming side, just needing time to transplant to the Web side? Or are Web advertisers proposing entirely new needs that require you to solve entirely new challenges?

A: When we entered this field, we also found that Meta ads have a very high share in the D2C field. Many advertisers naturally hope that other platforms can provide similar tools, such as precise exclusion functions, but we cannot, like Meta, precisely exclude existing users based on email or persistent identifiers, so there is a difference in targeting accuracy.

Additionally, we mainly deliver full-screen ads and combine dynamic product ads, which can stimulate stronger purchase intentions. Most of our conversions occur within 1-2 hours of ad display, while Meta usually has a longer attribution window.

Although many advertisers currently hope for more familiar functions, in fact, with existing functions alone, our approximately 600 customers have achieved $1 billion in annualized revenue, which is less than 1% market penetration. So, despite lacking some functions, the product effect is already very considerable.

We believe advertisers will eventually realize that different platforms have different advantages, which is a good thing. At the same time, we insist on full process automation, not allowing customers to manually target, just input goals and budgets, and the system will automatically optimize. This is also the trend of advertising technology—using AI to automatically optimize conversions. We have already proven this in the gaming field and will continue to promote intelligent automation in the e-commerce field to enhance effectiveness.

Q: Can it be understood that you decided to expand openness because of significant performance improvement? How is the overall performance now? How would you evaluate the current product effect?

A: The series of optimization tasks mentioned in the last call, originally expected to take several quarters to complete, were basically achieved by our team within one quarter, such as the launch of the Shopify application, achieving one-click integration; dynamic product ads have also been deployed, effectively improving ad conversion efficiency; at the same time, deep integration with attribution companies has been completed, aligning the data caliber of advertisers and models. Additionally, we have been continuously optimizing the underlying model.

From the current growth rate and feedback from advertisers, the performance is already very strong, so we are now confident to open the platform faster.

Q: As you gradually open the self-service platform and onboard more customers (including new industries and different scale customers), have there been any significant changes in the scale or industry distribution of new customers recently?

A: As the platform gradually opens, customer types and scales will become more diverse. Previously, we had strict restrictions on the GMV scale of onboarding, ranging from $25 million to $100 million. The long-term goal is for any type and size of small business to use our platform—as long as this can be achieved, the platform will have a significant impact on the global economy and employment. Our current product can already support efficient onboarding and achieving expected results for different scale customers. If the referral stage can verify this, we will quickly fully open and accelerate expansion.

Q: Regarding the referral system, will this model affect operating costs or profit margins? Will old customers who refer new customers receive rewards or subsidies?

A: For referral incentives, we believe that every dollar of incremental value brought is very important, but the platform is so large that even if there are referral rewards, the impact on profit margins can be ignored. At this stage, customers being able to invite peers during a limited period is already a benefit, so we do not plan to provide additional subsidies.

Q: In the long term, how do you consider expanding ad supply, especially outside the gaming field? Is there a possibility for MAX to logically expand into non-gaming ecosystems in the future? If so, which directions will be prioritized?

A: In fact, if there is broader advertiser demand in the future, we will quickly consider expanding more traffic sources. There is every reason to access more types of media resources, such as large social platforms, music, news, sports apps, and even websites. Our target audience is not only playing games, they spend a lot of time on other apps. If we can continuously reach these users outside of large platforms, it will be very valuable for us and advertisers. So, this is not a distant plan, it may be promoted within the next one or two quarters.

Q: After the Apple vs. Epic case, has there been any change in overall spending on user acquisition (UA) by game companies? Have you actually felt the growth brought by related benefits?

A: Currently, there is no obvious change, this process will be slower than expected. Although some apps have tried to bypass the App Store to reduce commissions, but leading game companies are still cautious, usually spending several quarters optimizing the experience. When large companies truly follow suit, small and medium-sized manufacturers will accelerate imitation, and then our platform will significantly benefit. It is expected to start seeing impacts in 2-4 quarters, with more significant impacts in 4-8 quarters, especially reflected in ad prices.

Q: Now that there is no Apps business, free cash flow profit margin exceeds 60%. How do you consider future cash usage and capital allocation?

A: In the future, our capital allocation will continue the previous approach. First, we will focus on investing in our own business growth, such as recruiting excellent engineering and business talents to drive organic growth. Secondly, we will continue to return value to shareholders through stock repurchases.

Q: In the short term, how do you view the comparison between the ideal target customer scale in the US market and the international market?

A: It can be divided into two parts: one is customer type, such as local enterprises in Japan and domestic enterprises in the United States; the other is traffic source. We have mainly relied on word-of-mouth growth in the past, and it is actually difficult to acquire local customers, but it is easier to attract international advertisers. Currently, about half of the revenue comes from the US and half from overseas (we basically do not operate in China).

Although it is difficult to double revenue in the short term, such as Western companies finding it hard to enter localized markets like Japan and Korea, some overseas markets like Europe are similar to the US, with great potential. When we open more markets, growth outside the US will be very noticeable.

At the same time, we will also actively strive for local customers in Japan and Korea, and as the penetration rate in these markets increases, the benefits will be very considerable for us, and we have the ability to achieve global coverage.

Q: Will the speed of international customer onboarding be as fast as the rapid growth in the US in the fourth quarter last year, or will it be more stable and gradually promoted?

A: In fact, we have not specifically counted the source countries of customers. As the platform opens, gaming customers themselves are very globalized, and when they invite peers, peers will be distributed worldwide; while e-commerce customers are mainly in the US, the impact will be more concentrated in Western markets. So international customer growth will be globally dispersed, not concentrated in a specific region, and there will be no restrictions.

Q: In serving different types of e-commerce advertisers and meeting diverse bidding goals and purchase cycles, what experiences and discoveries have you made? What insights do these provide for the development of self-service advertising tools?

A: Compared to the mobile application ecosystem, e-commerce advertisers' attribution and system integration are more fragmented and complex. The mobile side only has two major MMPs (we have one), and attribution models and integration methods are very unified; but the e-commerce field is very fragmented, advertisers are accustomed to Meta's way, and integration and attribution standards are inconsistent.

In the past one or two quarters, we mainly tackled platform integration and compatibility with major attribution companies, ensuring advertisers can integrate with one click, see the data they need, and the model can efficiently optimize. Now these difficulties have been overcome, customer feedback is positive, and it has strengthened our confidence to further open the platform.

Q: In the core gaming business this quarter, are there any noteworthy model optimizations or performance improvements?

A: This quarter, the core gaming business did not experience double-digit leaps, but overall data continues to grow, and Q2, although generally seasonal, still maintained healthy growth. The model continues to be slightly optimized, advertiser returns remain strong, and the willingness to reinvest is high.

Although the next "major model upgrade" has not yet been released, business performance remains strong. Subsequent new categories, new demands, and technical optimizations of the game itself are expected to bring more increments, and there may be unexpected major improvements.

Q: As the e-commerce business expands, will there be changes in marginal profit margins? For example, after marketing inputs to acquire advertisers or accessing new traffic channels, can the business still maintain a high marginal profit of 80%-100%?

A: For accessing new traffic channels, revenue will still be recorded as net income, not affecting overall profit margins; while new marketing inputs will be reflected in cost items. We have always managed marketing expenses very prudently, only investing when it is certain that these inputs can bring considerable returns. The growth of this part of marketing expenses precisely indicates that we are efficiently acquiring customers, with high investment returns (LTV/customer acquisition cost), which is good news for shareholders.

Q: Meta's re-launch of advanced mobile attribution, does this provide an opportunity for you to showcase incremental effects of application-side engines and e-commerce ads? Can you compare Meta's performance? Can advertisers participating in this plan help you gain greater attraction among e-commerce advertisers, especially after the invitation-based promotion goes live in October? Does this promotion pace help maintain the 10% target for annual e-commerce revenue?

A: Regarding Meta's advanced mobile attribution, we do not comment on the attribution integration of other companies. Meta has always been integrated with Adjust as a self-attribution channel (Adjust and Aspire are similar, operating independently), and we have no overlap with Meta in this regard, nor direct association.

Regarding e-commerce revenue share, it is currently about 10%, with market penetration less than 1%. Meta has over 10 million advertisers, and our business scale and opportunities are far greater than gaming. As long as performance continues to improve, the e-commerce business will soon become more important, and the future share is expected to continue to rise. Meanwhile, the gaming business will also continue to grow, both businesses have huge development space.

Q: The performance guidance for this quarter is more optimistic than last quarter, with the sequential increase rising from 4% to 5%. Is this mainly due to the return of user acquisition (UA) spending from some studios reflected in revenue? Is this understanding accurate?

A: The performance guidance we provide is mainly based on the most confident and predictable growth parts, such as continuous optimization of technology and model reinforcement learning, which generally contribute 3%-5% growth each quarter. This quarter's guidance is higher than usual, mainly due to the additional revenue from the divestiture of application assets.

Q: After the self-service platform goes live on October 1 with an invitation system, how many advertisers are currently on the waiting list? How many of these waiting advertisers have reached the GMP threshold, or have not been able to onboard due to capacity constraints? What is the scale of this identifiable, pending demand?

A: The invitation system is designed to allow successful existing advertisers to recommend new customers. Previously, our team manually screened, with very strict restrictions, and now through advertiser recommendations, the scale of new advertiser introduction will be much larger. But still adopting "invitation system," not fully open all at once.

Q4 will be different from the past, a new batch of advertisers will fully participate, driving business benchmarks to rise. Subsequently, it will be pushed to the international market, and when the platform is fully open in the future, it is expected to usher in a new round of growth and achieve more long-term stable development.

Q: Will the self-service ad management platform launched on October 1 directly include intelligent agent-related functions and automated optimization? Or will the initial version mainly have basic functions like credit card payment and performance attribution, with more advanced AI capabilities gradually launched later?

A: When launched on October 1, it will carry some AI-assisted functions, such as basic AI "assistants" to help advertisers with inquiries and guide usage, these tasks are relatively simple and will be launched early.

More complex AI functions, such as automatic analysis of ad effects and intelligent agents directly interacting with advertisers, are still under development and will be gradually introduced later. Additionally, there are generative AI ad creative tools that will support advertisers in automatically generating diverse video or card ads, making ad formats more diverse and enhancing conversion, especially helping small and medium customers improve capabilities. These functions will be gradually introduced, and the platform will continue to iterate, constantly introducing new tools to help advertisers use the advertising platform more efficiently and automatically, benefiting both consumers and us.

<End here>

Risk Disclosure and Statement of this Article:Dolphin Research Disclaimer and General Disclosure