Applovin (Minutes): Customers are waiting in line to join

The following is the minutes of the conference call for the Q1 FY25 financial report of $AppLovin(APP.US) For the financial report interpretation, please refer to "Short interest rotation? Can't withstand the impressive results."

I. Core Financial Information Review

1. Overall Performance:

a. Revenue & Profit: Total revenue increased by 40% year-on-year to $1.5 billion, adjusted EBITDA grew by 83% to $1 billion, with a profit margin of 68%, up 600 basis points from the previous quarter, demonstrating technology-driven revenue growth and cost control capabilities.

b. Cash Flow: Free cash flow of $826 million, up 113% year-on-year and 19% quarter-on-quarter, with an EBITDA to free cash flow conversion rate of 82%; cash and cash equivalents at the end of the quarter were $551 million.

c. Share Management: Repurchased 3.4 million shares at a cost of $1.2 billion, funded by free cash flow and credit (already repaid), reducing the number of outstanding shares to 338 million at the end of the quarter.

2. Advertising Business: Revenue of $1.16 billion, adjusted EBITDA of $943 million, with a profit margin of 81%. AI technology optimization improved advertising effectiveness, and online advertising solutions remained strong, combined with seasonal high spending in e-commerce. Revenue to EBITDA conversion rate quarter-on-quarter was 104% (adjusted 100%), maintaining efficiency despite increased data center costs.

3. Business Strategy Adjustment: Signed a contract with Triple Dot Studios to divest the mobile gaming business for $400 million in cash and 20% equity in the merged business, expected to complete in Q2, focusing on the advertising core business.

4. Q2 Performance Guidance: Only for the advertising business, expected Q2 2025 revenue of $1.195 billion - $1.215 billion, adjusted EBITDA of $970 million - $990 million, with a target profit margin of 81%.

II. Detailed Content of the Financial Report Conference Call

2.1 Key Information from Executive Statements

1. Q1 Business Performance: In Q1 2025, overcame seasonal challenges in the advertising industry to achieve significant growth. Optimized machine learning models helped mobile gaming companies expand advertising, while online advertisers contributed significantly throughout the quarter, with business diversification bringing positive effects.

2. Business Adjustment: Announced the complete sale of the gaming business, focusing on the core advertising business, and thanked the gaming studio team for their contributions in technology building, supporting their transition to game development.

3. Three strategic priorities for 2025:

a. Strengthen machine learning: Utilize AI to rapidly develop and continuously optimize models, creating greater value for partners and consolidating the platform's leading performance.

b. Advance e-commerce and online advertising solutions: Focus on three areas, continuously improve models; although there have been results, it is still in the early stages of iterative optimization. Strengthen integration with third-party platforms and attribution providers; due to the fragmented online advertising market, it will take time to build a seamless measurement experience; develop a self-service dashboard, which will be launched to some customers this quarter, and after full rollout, will enable automated ad placements. Currently, the pilot results for online advertising are good, but market penetration is less than 0.1%. New partners will drive growth, and after the tools are improved, global promotion will unleash huge opportunities.

c. Optimize ad creativity: Enhance ad testing and automated creation to improve creative experiences, helping advertisers easily optimize their ad campaigns.

4. Tariffs: Over 90% of advertising revenue comes from mobile games, which are not directly affected by tariffs. Online advertising focuses on the mid-market; although some partner merchants may be affected by tariffs, the low market penetration means that tariffs have no significant impact on business development.

5. External dynamics: AppLovin welcomes competition and is confident in maintaining its lead due to its data and AI advantages. It is optimistic about the potential alternative payment systems in the Apple App Store, as their low fees benefit customers and encourage increased user acquisition spending, which in turn benefits the platform. Details on TikTok's bidding can be found in the blog.

2.2 Q&A

Q: The guidance for next quarter shows a sequential growth of 3% - 5% in advertising, which seems slower than the previous long-term guidance for mobile advertising. What is the reason for this?

A: The business is significantly affected by seasonality. The first quarter is usually the worst due to holidays, Ramadan, and other factors; the second quarter has no special benefits; the third quarter is summer, and the fourth quarter has holidays. Last year's sequential growth showed only the second quarter had single-digit growth. We expect traditional mobile advertising, such as games, to grow 20%-30% year-on-year, but the actual may be higher, as the first quarter saw nearly 20% sequential growth compared to the fourth quarter. Currently, advertisers are very enthusiastic, with spending on games and online advertising reaching new highs.

Additionally, this quarter we will launch a new dashboard to some advertisers to collect feedback; after the full rollout of self-service in the future, the platform will significantly expand its user base, leading to business transformation. Past data cannot reflect future trends, as the business has not yet entered a stable development phase.

Q: I have noticed that in some past quarters, advertising revenue has declined sequentially, but this has not affected long-term growth. Should investors expect similar sequential declines in the future, and that such declines will not shake the fundamentals of the business? I recall that this happened in 2022.

A: 2022 is very different from now; the AXON model had not yet been launched then. Since its launch, the business has achieved significant growth through the flywheel effect of machine learning—more impressions, interactions, and conversions make the model smarter, coupled with the team's continuous technological advantages, leading to rapid development.

Q: Last quarter, you emphasized new industries such as fintech, healthcare, and insurance, and have achieved success. I want to ask, aside from the industries mentioned, will you continue to expand into other new industries? Q: What pain points do they hope to address when communicating with new types of advertisers? Is it about connecting with attribution partners or issues related to self-service?

A: We are currently not in a hurry to expand into new online advertising areas. There are a large number of existing clients waiting to join the platform, and although we have signed hundreds of advertisers, our team is limited in size. We plan to proceed gradually, and for now, we need to launch self-service tools to achieve process automation. The online advertising model is still in its early stages and has significant room for improvement.

By strengthening attribution and platform integration and optimizing the model, the performance of the model will continue to improve as the team makes ongoing enhancements. We are confident that we can reach a leading level in online advertising comparable to game advertising. Although this has not yet been achieved, early performance has exceeded expectations. In the future, combining self-service dashboards with global promotion will strongly drive long-term growth.

Q: Is automated ad creation essentially customized creativity? If so, considering the cost of generating creative content for advertisers in real-time, how does the cost-benefit analysis look?

A: The cost-effectiveness cannot be determined until practical operations begin. We plan to create ad creatives based on generative AI. Currently, the platform serves over 1 billion daily active users, resulting in a large volume of ad displays. For example, if an advertiser uploads 20 static videos for placement, it can achieve a certain response rate. With the help of large language models and customization technology, while we cannot create completely different videos for each user, we can process high-quality videos uploaded by advertisers to dynamically generate more content, achieving deeper personalized ad recommendations on the platform.

Testing ad creatives is key to improving ad response rates, as it is low-cost and effective. If we can systematize this (we are very confident about this, but it will take time to bring it to market), it will significantly increase audience response rates. Each successful instance not only brings short-term growth but also accumulates over time—the system optimizes learning from more transactions and continues to reinforce itself. This is a highly impactful direction and a top priority for our subsequent work this year.

Q: What new insights do you have regarding advertiser churn? Is there a churn phenomenon, and what are your expectations for churn rates in the future? Does advertiser spending per account increase over time, or is it only influenced by seasonal factors?

A: We previously disclosed that the platform has over 600 e-commerce advertisers with an annual revenue of $1 billion. Due to the non-comparability of sequential data in the shopping sector between the first and fourth quarters, we will not disclose the growth in the number of advertisers for now. The online advertising business is still in its early stages, and customer churn is unavoidable. Recent statistics show that the churn rate for advertisers spending over $250,000 annually is below 3%, but this still does not meet our expectations.

In the game advertising sector, advertisers rarely leave unless the game completely fails, as we are key to their success. We hope to achieve the same in the online advertising space. The online advertising product has only been launched for a few months, and as we optimize the model, we are confident that we can become the marketing platform of choice for all advertisers with websites or applications, helping them achieve success.

Q: You disclosed the year-on-year growth of net income per installation. However, e-commerce advertisers do not bring installation volume but rather user behavior volume. Does this metric include user behavior data driven by e-commerce?

A: Not included. This metric covers the revenue of online advertisers, and since the installation metrics remain unchanged, it will increase the net revenue per installation, which is solely based on cost-per-install (CPI) advertiser data.

Q: You mentioned that some online advertisers still have room for improvement in their advertising effectiveness. I would like to know if these advertisers share any common characteristics? Looking ahead to the second quarter of this year and beyond, what progress or landmark achievements can we expect in improving their advertising effectiveness?

A: This is not a specific trend; the model does not operate that way. Looking back at past game advertising, whenever the model's predictive ability improves, client returns and scale also grow. Based on the current market deployment time of the online model, its performance is quite good; I would rate it a B+, but there is still significant room for improvement. The growth of the gaming business has been remarkable over the past few years, and since the launch of AXON 2, platform advertising spending has roughly quadrupled, thanks to higher return rates and larger scales brought by model improvements.

If engineers release a new version of the online model tomorrow, a 30% - 40% improvement in effectiveness would not be surprising. This does not mean that the return on advertising spend (ROAS) directly improves, but rather that advertising scale increases by 30% - 40% at the same ROAS, and this growth will continue to accumulate and spread on the platform.

Q: When communicating with advertisers, they generally hope to exclude specific audiences for their advertising campaigns. I would like to know if there have been any new developments regarding this since our last communication about online advertising services?

A: This service is quite interesting and relates to advertisers' spending habits on the Meta platform. Many advertisers wish to upload exclusion lists and target new audiences proportionally, rather than engaging in repetitive marketing. Our product is still in its early stages, and the key to measuring success is to scale up according to the advertisers' expected return on advertising spend, focusing on optimizing model matching capabilities rather than audience segmentation.

In the future, we may launch a feature to exclude specific audience targeting. There have been attempts in the past few months, but it is not a current priority. If product performance continues to improve, it is expected to enable hundreds of advertisers to achieve an annualized revenue of $1 billion, while the current market penetration rate is less than 0.1%. Although we are still in the early stages and cannot meet all advertisers' needs, opening the platform will lead to significant growth. We do not aim to imitate other companies' products but are committed to creating higher-quality products for advertisers. We will consider feedback but will still focus on improving the model to provide the best results.

Q: I would like to understand the speed of new online advertisers this quarter and how that speed has changed since the report of 600 clients in December. Additionally, what are the expectations for the second quarter and the second half of the year?

A: Due to limited resources, the speed of acquiring new online advertisers has slowed down. The team consists of only about 20 people, and there is a shortage of personnel in various positions. The current goal is to launch a self-service dashboard and automation tools to restore the speed of onboarding advertisers to the level achieved when we first reached 600 clients. The current number of clients has exceeded 600, but the lack of manpower makes it difficult to manually handle a large number of onboardings.

The new dashboard is currently being tested, and the feedback has been positive. After using it, advertisers can achieve operational automation, helping the team accelerate the onboarding speed. We will gradually roll out this dashboard, allowing advertisers to try it out this quarter, and will continue to open it up progressively, ultimately fully launching and promoting it globally, significantly improving the efficiency of advertiser onboarding.

Q: Many e-commerce advertisers have reported that online products significantly optimize the 24-hour conversion window for advertisers. I would like to know what progress has been made in assisting advertisers with long consideration cycles and high repurchase rates?

A: We do not have a conversion API and lack data such as emails and phone numbers, so we can only attribute conversions within the cookie validity period. Currently, browsers like Safari have a very short cookie validity period, which requires our model to quickly prompt user actions. This is not a problem for most advertisers, as most products are not high-priced.

If a $200 shirt is sold in a short time, advertisers can clearly see the effect of the advertisement; however, if the sale occurs two weeks later, it is difficult to prove that it is due to our platform. For advertisers selling high-priced products, since we cannot facilitate transactions in a short time, we may only be able to optimize the early stages of the funnel, such as phone communication or email registration.

Currently, we have hundreds of advertisers, and the market space is large, with most products selling quickly, so we have not yet considered optimizing high-priced product sales. There is still significant room for expansion in the future.

Q: I would like to ask about the self-service model. You are currently testing it and would like to know how advertisers are expected to react once the model is launched. Will they participate enthusiastically with large budgets, or will they test first and then gradually increase their budgets? I would like to hear your expectations for the situation after the launch.

A: Advertisers need to confirm that spending money is worthwhile; they will not initially pour a large budget into the new system. Typically, they will gradually increase their budgets after seeing that their advertising campaigns meet their expected goals over a few weeks or months.

Currently, we have limited the online advertising audience to the United States as a strategic arrangement. Our business is roughly split evenly between domestic and international (excluding China), and there are huge opportunities to expand into global markets. We will combine the self-service dashboard with our globalization strategy to quickly expand the advertising audience and increase the number and types of new advertisers. The subsequent performance will depend on how advertisers leverage it.

It is worth mentioning that we are the last large advertising company to launch a self-service dashboard; other social platforms, search platforms, and large display channels already have similar advertising management products. Once the product is launched, agencies and advertisers will naturally know how to use it.

Q: Previously, you mentioned that the growth algorithm was about 20% - 30%. Now that the business is growing rapidly, is this model description still applicable?

A: We still believe that 20% - 30% is a suitable long-term growth rate for the company. This growth consists of two parts: one is the model's reinforcement learning and self-improvement through continuous transactions, which accounts for about 3% - 5% and is also the quarterly guidance data, with strong stability and predictability; the other is targeted improvements, such as model iterations and upgrades. The engineering team’s introduction of new models can bring stepwise growth, and this part currently adds about 10% to the annual growth rate. In the future, there will be at least one such stepwise growth each year, and based on past performance, this expectation is already relatively conservative.

Q: Regarding the next-level plan, specifically involving user acquisition between non-gaming applications and the previously mentioned dynamic personalized advertising creatives, will these plans be able to take effect in 2025, or are they more likely to be part of the 2026 planning? A: The so-called "new business" is actually an expansion of existing business. This year, the focus is on implementing the plan list, and doing well in these areas will lay a solid foundation for development in 2026. We have been making steady progress in 2025. Currently, no other technology company can match us in terms of both strong financial health and rapid growth scale.

We are focused on development over the next two years. If we successfully launch the self-service dashboard, achieve media buying automation, and lower the barriers for advertisers, the business will continue to grow rapidly for many years to come. In the future, we will transform into a marketing-oriented company, attracting small businesses to advertise, establishing the LTV-CAC model, and leveraging a large potential advertiser base to unlock significant value.

In addition, we are making efforts in various areas such as dynamic ad creativity, model optimization, and system self-learning. These growth drivers give us confidence in both the present and the future.

Q: I remember you mentioned that the non-gaming audience business might account for over 10% of total advertising revenue this year. Is that still the expectation for this business in 2025?

A: Regarding online advertising and its contribution to revenue, we previously estimated its proportion to be around 10%. However, this year's proportion is difficult to predict due to various influencing factors.

The development of the mobile gaming business is related to model improvements. If the improvements are significant, its growth rate may exceed that of the e-commerce business. However, after launching the self-service model, we are very confident in the e-commerce business, which may see substantial growth, likely causing the proportion of online advertising revenue to exceed 10%.

Q: Last year, there were two quarters (Q1 and Q3) of artificially guided improvements. Is it still possible for such improvements to occur in the AXON 2 phase over the next three to five years? It depends on continuous learning rather than the stage of the machine, right?

A: The application of neural networks is still new, and model iterations are rapid, with many studies driving engineers to develop new versions. We are keeping up with the forefront, and the technology implementation is excellent, but the industry is still in its early stages. The team needs to conduct in-depth research, and the AXON model will continue to iterate in the future. Each major improvement can bring double-digit growth quarterly, and even small optimizations can significantly impact the business. We have a lot of work ahead, and the research team will continue to enhance model performance.

Q: Are the learning mechanisms for gaming and e-commerce businesses the same? Or can the e-commerce business make progress through learning within a quarter while the gaming business cannot?

A: Gaming and e-commerce are two independent models, both undergoing reinforcement learning and targeted improvements simultaneously. The same small team is responsible for both models, continuously testing potential optimization solutions to achieve stepwise growth. The two develop independently, and the e-commerce business may grow faster due to the addition of new advertisers.

Q: Is it possible that during the operation of AXON 2, the e-commerce model improves efficiency due to certain circumstances? They improve at their own stepwise pace, is that what you mean?

A: I always say that the e-commerce model is still in its early stages, and that is the reason. The AXON 2 model in the gaming field has undergone multiple iterations, giving the team more time to optimize. In contrast, the e-commerce model has only been running for a few quarters, and due to its short duration and insufficient display volume and transaction scale, it lacks a complete data feedback loop, making it difficult to retrain itself

Q: When reviewing Q1 performance, I want to understand if the approximately 20% quarter-over-quarter growth in the advertising business is due to a stepwise breakthrough rather than improvements from reinforcement learning? Has the quarter-over-quarter growth rate in the gaming business exceeded the typical low to mid-single-digit level?

A: Q1 had two fewer days, and although there were holidays, it was not as strong as the Q4 peak season. However, it was a quarter of significant growth, and we may be the fastest-growing tech company of our scale, driven by multiple factors.

I previously mentioned improvements in model efficiency; the team has been continuously optimizing the model. Although there have been no major changes, the results are significant, and self-learning is one aspect of this.

E-commerce advertising only had a full quarter of contribution in Q4 last year, but its impact is smaller than that of the gaming business. This quarter, over 50% of the incremental growth came from the gaming sector (Dolphin Research: implying a year-over-year growth of 35% in gaming advertising, etc.). Currently, there is room for improvement in both advertiser reliance and the conversion rate per thousand impressions. The technology is still in its early stages, and the team's subsequent work is ample. We are very optimistic about the development opportunities for the model under the flywheel effect of scaling up and continuous optimization.

Q: Does the Q2 performance guidance include revenue or recognized income from traditional studio expenditures? I remember there was a disclosure of the total value of software transactions. From an accounting perspective, these studios are considered third parties, and their income is not recorded. As long as they are still operating, there may be transitional situations. Are these situations reflected in the performance data? Is this currently important?

A: After signing the agreement with Triple Dot to divest the application business, the transaction is expected to be completed by the end of this quarter. The performance guidance does not include the incremental revenue brought by the studios transitioning to external parties, nor the premium on user acquisition costs. Subsequent expenditures will depend on Triple Dot's operational situation, which overall does not have a significant impact on the business. Although there are additional revenues, it is difficult to significantly change the business situation.

Q: Can you elaborate on the regulatory dynamics of app stores and the potential fee reductions for mobile games this year? I want to understand: first, what is the potential impact on the business from the perspective of advertising expenditure? Second, how to layout the business to seize the additional advertising expenditure opportunities brought by fee reductions?

A: We are currently the largest channel for mobile game user spending and the best advertising platform, likely to become the biggest beneficiary of this change. If app store fees are halved, mobile game developers' income will increase. For example, if they originally received $0.70 per income, with the fee reduction, they could receive $0.85, an increase of 20%.

Developers will invest the increased income into advertising. Since we use a dynamic auction mechanism, once a developer increases their bid due to increased income, other developers will follow suit to compete. Over time, this change will benefit the entire ecosystem and drive growth. More funds will flow into the market, and advertisers' bids will become more competitive. Both we and the publishers will benefit from this, ultimately enhancing the company's profitability and benefiting shareholders.

Q: How is the ability to monetize profits based on a relatively fixed inventory base? The e-commerce business has been launched for several quarters; I want to understand your views on this and whether the situation has changed?

A: We have a large fixed inventory base, similar to Meta and YouTube, with limited new user growth, especially in the U.S. market. However, well-performing performance advertising companies are strengthening their technical algorithm matching capabilities, while we started at a lower level In the past, we mentioned that the conversion rate for facilitating transactions was 1%, which has now significantly increased, and this figure is expected to continue to rise.

Without a substantial increase, our business can still achieve expansion, and the future conversion rate is expected to rise to 2% - 5%, with huge potential. We focus on full-screen videos that can attract users' attention. As the self-service platform opens and attracts a large number of customers, more content will be displayed to consumers, along with personalized advertising creativity. Enhanced consumer feedback is not only beneficial to us but also helps publishers, which is expected to drive the company towards long-term growth.

Q: A small question about the macro situation of games. From the performance, there are no signs of weakness. Are there any noteworthy aspects regarding the release of new games from studios or performance in various regions?

A: Currently, our business is highly diversified, and a single game is unlikely to have a significant impact on our operational scale. Large-scale releases and marketing activities have been conducted around several games in the ecosystem, but there is no difference in the rhythm of game releases compared to the past. Tariffs will not affect the digital economy; free games, as a cheap and easily accessible form of entertainment, are less affected by economic fluctuations, and we are confident in our market position.

Q: There are two questions regarding the self-service dashboard. First, who will have priority access to this dashboard? Is it currently aimed at existing customers, or is it open to all customers regardless of annual budget?

A: We will roll out the self-service dashboard in phases. Initially, it will be available to a feedback group composed of existing customers, and then it will be opened to all existing customers, which will alleviate the team's manpower burden and will be advanced in the coming weeks. In the short term, we will allow customers to use the dashboard, and in the medium to long term, it will be opened in phases. It will be fully opened after several quarters, gradually opening to new customer types during this period to ensure customer quality, ensure the platform operates without issues, adapt to various scenarios, and allow time for model optimization and improvement.

Q: Considering its potential in providing more data and enhancing processing capabilities, will this help improve performance in the short term?

A: The more customers there are, the better the recommendation model becomes. Just like beauty companies entering the platform, having 1,000 companies can provide consumers with a more diverse range of products than just 10, and more recommendation options can enhance model performance. We are currently not perfect; we cannot adapt products to all customers like we can in the gaming field.

We cannot have situations where some do not operate normally when a large number of customers enter. Once the platform opens and attracts many customers, the business scale will grow exponentially, with astonishing growth rates. However, we pursue perfection, hoping that the product can perfectly adapt to all customers and create the best-performing product. Although we have not achieved this yet, opening the platform after reaching our goals will drive long-term growth.

Q: I read the blog about TikTok. I understand you said the opportunity is "slim," but it seems you are more focused on the Trump administration's restrictions on TikTok's U.S. operations, and this proposal seems to cover a broader scope than what the U.S. side is considering, targeting globally (excluding China). I want to confirm the message conveyed in the blog and how you view Washington's related actions and your positioning.

A: In short, we are targeting markets outside of China, and our top priority is to address U.S. national security concerns regarding TikTok's algorithms and data, which is equally critical in other regions. To achieve this goal, we must have operational control and need a company that can rewrite parts of the algorithms to ensure compliance with security standards We believe our solution is the most feasible, relying on model knowledge and proposals, perhaps the only one that can achieve this. Through the proposed cooperation model, we can solve current problems and ensure long-term operations. Although the hope is slim, we dare to challenge. We have top-notch performance advertising AI models, combined with TikTok's large user base, which is expected to create significant value for shareholders. Details have been published on the blog.

Q: You are improving efficiency while experiencing rapid growth. What new insights do you have regarding medium to long-term marginal profit margins and stability levels?

A: We will continue to operate the business in line with the company's philosophy, focusing on efficiency while growing revenue. The self-service platform is a way to achieve growth without increasing costs. It is expected that, aside from data center costs (which are highly variable), major costs will remain stable.

Currently, the annual growth rate of data center costs accounts for about 10% of revenue growth and will continue. The profit margin of the advertising business will continue to improve from the current level, and we will see how far it can go.

Q: Can you analyze the development trends of application products and web products and distinguish between the two? The boundaries are becoming increasingly blurred.

A: Essentially, this is about marketing content through our platform, relying on models to produce results. Whether the advertisement is for an app, a website, or an in-app webpage, the model should be effective. Initially, we focused on mobile game ads, and the first business version in 12 years concentrated on a single use case, which has now expanded to web and combined use cases.

In the long run, we envision advertisers informing us of their targeting goals, products, expected returns, budgets, and uploading videos, after which AppLovin can handle the rest, achieving personalized advertising and producing satisfactory results. In the future, these business contents will ultimately merge.

Q: You mentioned that the web business is expected to contribute 10% this year. Does this refer to gross profit or net profit? Additionally, you handle a large amount of advertising inventory; how do you distinguish the differences?

A: The 10% contribution mentioned here is calculated based on net revenue. We believe that web advertising solutions can contribute at least 10% to the overall net revenue of the advertising business.

Q: Regarding the issue of self-attribution for long-term development. Currently, there are shortcomings in this area, such as a lack of identity recognition and other key elements. I would like to know if you will always rely on third parties, or will you be limited to low-attention, high-turnover products? Or, with the existing huge demand and future growth potential, will you gradually tackle the challenges? Can you share your thoughts?

A: It is indeed the third situation; we have a large amount of demand and opportunity. On the app side, we rely on third-party attribution like Appsfire and Adjust, while web products can achieve self-attribution on our own platform, and they are not high-turnover products. Our model can generate value within minutes after an ad is displayed. Considering a $1 billion operating scale, although time is tight, it also drives us to operate efficiently, allowing advertising investments to yield quick results and reducing doubts about attribution.

In the long run, we hope advertisers can measure us according to their own needs. Given that different companies use different attribution tools, we will integrate third-party solutions on the web side just as we do on the app side, providing advertisers with multiple reference dimensions to reassure them to increase their investments. Our goal is to become the number one or number two channel for advertisers, earning their full trust in their funding investments

Risk Disclosure and Disclaimer for this Article: Dolphin Investment Research Disclaimer and General Disclosure