[IN-DEPTH ANALYSIS] AppLovin: The Secret to Explosive Growth Through AI and Acquisitions

Key Takeaways

• Platform Transformation: Through key acquisitions (MAX, Adjust, MoPub), AppLovin evolved from a mobile game ad platform to an industry-wide platform, expanding its technology, user base, and market reach.
• AI-Driven Breakthrough: Axon 2.0, powered by deep learning and large-scale user data integration, improved ad efficiency, particularly in e-commerce, driving rapid advertising revenue growth.
• Valuation Risks: Higher than average valuation (EV/EBITDA 46x) reflects growth, but an economic downturn could cut ad budgets, slow revenue, and lower valuation to 30-40x.

Source: Tradingview.com

In the digital advertising market, Google and Meta are undoubtedly the undisputed overlords. In 2024, the global advertising revenue of these two giants reach $265 billion and $160 billion respectively, accounting for about 50% of the market share, while AppLovin(APP)'s market share is only about 3%, which seems insignificant in the shadow of the giants. However, AppLovin did not choose to confront Google and Meta head-on, but cleverly found a niche market of its own, the mobile game advertising market. With 1.4 billion daily active users (DAU), most of whom are game players, AppLovin has deeply cultivated this niche covered by Google and Meta but not deeply explored, serving game developers and player groups. Axon 2.0, launched in 2023, further strengthened its competitiveness. Through AI technology, it accurately locates users who will spend money, not only consolidating the advantages of game advertising, but also successfully testing the e-commerce market. AppLovin's outstanding performance in 2024, revenue of $4.7 billion, a YoY increase of 43%, is not achieved overnight, but is the result of a series of precise strategic layouts in the past few years, especially several key acquisitions. These acquisitions not only injected technology and resources into AppLovin but also laid the foundation for it to move from a small giant in game advertising to an advertising platform for the entire industry.

Source: AppLovin

Three Important Acquisitions

AppLovin's business portfolio can be understood as a complete set of tools and services it provides to advertisers, with the goal of allowing advertisers to earn like $1.5 or even more in return for every $1 spent on advertising. Initially, it was just a small platform focusing on mobile game advertising, but through acquisitions, it has gradually built a powerful advertising ecosystem. These acquisitions are not just about buying tools, but strategically changing AppLovin's business model, market positioning and profitability.

MAX (acquired in 2018)

MAX is a startup founded in 2017 that focuses on in-app bidding technology. Simply put, it is a tool that helps app developers sell ad space to advertisers through real-time bidding, ensuring that each ad display can be sold at the highest price.

Before 2018, AppLovin was mainly an ad network (demand-side platform) that helped advertisers buy ad space, but it lacked an efficient way to help developers sell ad space. The traditional waterfall model is inefficient, and developers can only try different ad networks in sequence and often cannot sell at a good price. The waterfall model works like this: Ad networks are ranked based on historical performance data such as eCPM (effective cost per mille), fill rate, and latency. Ad inventory is offered to ad networks in a predetermined order, starting with the highest-ranked network. If the top-ranked network cannot fill the ad inventory, the request falls down to the next network in line, and so on, until the inventory is sold. It does not have simultaneous bidding which allows for greater competition among advertisers.

MAX brings in-app bidding technology, allowing all advertisers to bid at the same time, so that developers can get the highest price for each ad display. In the past, an ad space worth $1 might only sell for$ 0.8. Now with MAX, all advertisers bid together ($1, $1.2, $1.5), and developers directly get $1.5, and their income increases immediately. Advertisers can also spend $1.5 to reach more accurate users and earn $2, which satisfies both parties. AppLovin has upgraded from only helping buyers to helping buyers and sellers. MAX fills the technical shortcomings of AppLovin, allowing it to directly serve developers (supply side), which is a key step in the transformation to platformization and complements the later acquisition of MoPub.

Source: Naavik

The increase in the price of ad space (CPM, cost per thousand impressions) allows developers to earn more, and AppLovin as a platform also takes more commissions, which directly improves the monetization efficiency of the AppLovin platform. After developers earn more, they will invest the money in AppLovin's user acquisition tool (AppDiscovery), forming a revenue cycle. In 2018, in-app bidding was a new trend in the advertising industry, but mobile adoption was slow. AppLovin seized this key bidding opportunity and grabbed market share.

Adjust (acquired in 2021)

In the advertising industry, advertisers spend money on advertising, and what they care most about is the effect: how much can they earn back for spending $1. In the field of mobile advertising, tracking the effect is a difficult problem. This is because advertising involves multiple links (display, click, download, purchase), and the data of these links are often scattered or opaque, making it difficult for advertisers to know whether the money is worth spending. This situation is like a black box: you put $1 in, but you don’t know what comes out.

Before acquiring Adjust, AppLovin was mainly responsible for helping advertisers to invest $1 and display it on the ad space of its games or partner applications, but it could not directly track the subsequent results brought by the advertisement, such as whether the user downloaded the application or spent money. A game developer spends $1 to promote a new game but may only know that 100 people clicked on the advertisement, but it is not clear whether 10 of the 100 people downloaded the game, or 50 people downloaded it but no one spent money on props. Mobile advertising attribution, that is, to determine which advertisement brought a certain user behavior (download, purchase) requires cross-platform tracking capabilities. AppLovin does not have this technology itself and can only rely on rough feedback from advertisers or apps. Users may see the promotion in AppLovin's ads, click to download, but spend money to buy things a few days later. The chain in the middle is broken, and AppLovin cannot tell advertisers that it earned this money because of this ad. Therefore, advertisers are unsure about the effectiveness of AppLovin. They may feel that spending $1 seems to have some effect, but they are not sure whether it is worth it, so they dare not invest more money. This uncertainty makes advertisers tend to give their budgets to platforms with more data (such as Meta, which can clearly report how much money is earned for spending $1). AppLovin's order volume and revenue growth are therefore limited.

Adjust is a mobile attribution and analysis tool that helps advertisers track the effect of every dollar of advertising fees, such as which channel it is spent on, how many users it brings, and how much return it finally earns. Adjust provides detailed data (for example, this $1 brings 10 new users, 2 of which spend money), which reassures advertisers and completely changes the advertiser experience.

Full link tracking: Adjust embeds tracking links and software development kits (SDKs) into ads and apps to record every step from when the user sees the ad to when the user finally takes action. When a user sees an AppLovin ad and clicks it to download it from the app store, Adjust records that this is a click from the ad. When the user downloads the app and opens it, Adjust marks this as an installation from the ad. When the user spends money in the app (such as buying props or goods), Adjust captures this as revenue from the ad through the SDK.

Precise attribution: Adjust uses fingerprinting and device matching technology to link user behavior to ads. When a user clicks on an ad but does not download it immediately, Adjust uses the device ID or IP address to confirm that the download is from the ad within a certain time window (usually 7 days). Even if there is a delay in the middle, such as buying something the next day, Adjust can trace it back to the initial $1 ad.

Source: AppLovin

The data collected by Adjust can also be fed to AppLovin's AI (Axon), making advertising more accurate, and they may earn $2 for spending $1 next time. At the same time, with detailed data, advertisers can adjust their delivery. For example, if Adjust shows that a certain area only earns $0.5, they will stop the advertising there and invest $1 in the place that earns $1.8. This ability makes advertisers feel that AppLovin is more reliable.

Adjust is the first step for AppLovin to shift from an extensive advertising network to a refined service platform. In 2023, the proportion of advertising revenue increased from 40% to 60%, and Adjust contributed greatly. It improved advertiser return rate and ARPU (revenue per user), directly pushing up the company's valuation.

MoPub (acquired in 2022)

In the advertising industry, inventory refers to places where ads can be displayed, such as banners, videos, or interstitials in games. Advertisers spend $1 to buy this inventory, hoping to reach users and eventually earn $1.5 back (for example, through users downloading apps or purchasing goods). The number and quality of inventory directly determine the scale of users reached and the return on investment (ROI).

Before acquiring MoPub, AppLovin's inventory mainly came from its own game assets (such as games developed through acquisitions such as Machine Zone and Lion Studios). Although AppLovin owns multiple game studios, the inventory it can generate depends on the user scale of these games. Its games may have tens of millions of daily active users (DAU), which is far lower than industry giants (such as Meta's 3 billion DAU). With few inventories, advertisers can only reach millions of users. Moreover, the users of its own games are mainly gamers (such as young people aged 18-35), and their interest is focused on in-game purchases or related products. Other types of advertisers (such as e-commerce companies selling clothes or cosmetics) may feel that these users do not match their target customers. Spending $1 on advertising may reach the wrong people, and the probability of earning $1.5 is reduced. For example, a cosmetics advertisement placed on a male user who only plays games may only earn $0.5.

AppLovin had to develop and operate games to generate advertising space, which is costly. It may cost millions of dollars to develop games, and the life cycle of games is limited. Popular games may become outdated in a few years. The supply of advertising space is unstable, and the effect of spending money may be good and bad at times, so advertisers are unwilling to invest in the long term. So, in 2022, AppLovin acquired MoPub from Twitter for $1.05 billion. MoPub is a supply-side platform (SSP), which is equivalent to an advertising trading market. It sells advertising space of a large number of applications to advertisers, and AppLovin takes a commission from it.

Reach more users: MoPub served tens of thousands of applications (including news, tools, life, etc.) at the time, covering 700 million daily active users. After the acquisition, AppLovin integrated these ad spaces into its own platform, and the inventory scale expanded by more than 10 times, and the number of people reached increased by 70 times. The more users there are, the greater the chance of seeing the ads. The advertising industry has a scale effect, the larger the inventory, the lower the cost of a single delivery and the higher the return.

Reach the right users: MoPub's ad spaces come from various applications, not just games. For example, users may read articles on news apps and record exercise on fitness apps. These groups include women, middle-aged people and other groups other than game players. Ads need to be seen by the right people. MoPub's diversified inventory has transformed AppLovin from a game advertising expert to an advertising platform for the entire industry, broadening its customer base.

Improve stability: As an SSP, MoPub aggregates ad spaces from external applications, and AppLovin can obtain inventory without developing games itself. It's like going from growing vegetables to opening a supermarket, with a constant supply.

This reduces AppLovin's operating costs, and it doesn't have to spend millions to develop games, and its profit margins are higher. The SSP model is the trend in the advertising industry. MoPub allows AppLovin to shift from content-driven to platform-driven, improving both efficiency and stability.

AI: Driving Business to A Higher Level

In the mid-2010s, AI technology emerged, especially in the advertising field, such as Google's DoubleClick using AI to optimize bidding. AppLovin realized that in order to compete with big players, it must replace manual labor with AI, so it invested resources to develop Axon. It analyzes user data to predict who will click on ads, download apps or buy goods, and then accurately deliver ads to them. Axon is based on machine learning and relies on a large amount of user data for training, AppLovin's early game advertising experience and data, such as which players are willing to spend money on props.

Axon 1.0 was launched in 2018 and is AppLovin's first AI advertising engine. It mainly serves game advertising and relies on the gaming habits (duration, clicks, in-app purchases) of tens of millions of DAU of AppLovin's own games. Axon 1.0 is like a game expert, but it is confused when it leaves the game circle. The predictions are not smart enough and the data is not wide enough.

Axon 2.0 was launched in 2023. It is a comprehensive upgrade of Axon 1.0, transforming from games to e-commerce and CTV. Compared with 1.0, Axon 2.0 has achieved significant progress in multiple dimensions: wider coverage (from a single game to multiple industries), stronger data processing capabilities (supporting big data analysis), smarter algorithms (using deep learning technology), and faster optimization speed (real-time adjustment).

Source: TradingKey

In terms of data support, Axon 2.0 has greatly expanded its sources and scale. It no longer relies solely on users of AppLovin's own games, but integrates DAU brought by MoPub and Wurl's CTV user data, covering richer user behavior scenarios, such as browsing news, watching videos, switching applications, etc. The amount of data has jumped from hundreds of millions to trillions of requests, and the types are more diverse, including multi-dimensional information such as clicks, duration, and geographic location. This upgrade allows Axon 2. 0 to handle complex needs that were previously unattainable. At the algorithm level, it has evolved from simple machine learning in 1.0 to deep learning, using more complex models for analysis. Axon 2.0 can not only recognize that this person plays a lot of games but also infer that he plays fewer games but watches videos frequently and may be more inclined to buy electronic products or subscribe to services, thereby achieving more accurate advertising. This ability enables it to effectively predict user intentions under privacy restrictions, significantly improving advertising effectiveness. It has improved ad click-through rates by 15% and conversion rates by 20%.

Difficulty: Replicating Success of Gaming to E-Commerce

The closed loop formed in the game field is not so smoothly realized in the e-commerce field, because in the game field, developers are both advertisers and publishers. Game developers use AppLovin's platform to promote their own games and sell advertising space in their own games to other developers. DAU is both a goal and a monetization tool.

However, the goal of e-commerce advertisers is not DAU, but GMV. They don't just want to increase the number of users but want to sell goods to make money. They spend $1 to directly earn $1.5 in sales, rather than increasing DAU and then relying on advertising monetization. User accuracy is more important than quantity. If the users who are attracted do not buy things, it is not of much value to the merchant but increases operating costs. Therefore, AppLovin has no way to directly copy the successful model in gaming.

The problem here is that AppLovin has data in the game field but does not have a large amount of e-commerce user data. Unlike Meta, which has social and shopping records, Meta can accurately find people who will buy things by matching users' social, interest, purchase history and other data. In the short-selling report released by Fuzzy Panda Research and Culper Research, it can be seen that in order to solve this shortcoming, AppLovin may analyze the advertising data on the Meta platform, reversely deduce Meta's user portrait, and optimize its own advertising. Short sellers believe that AppLovin's success is not based on self-developed technology but stealing Meta's data through improper means.

While fact, Meta can't do much to prohibit this behavior at this stage, because it is authorized by advertisers to AppLovin. Advertisers are not loyal to a single platform and only care about ROI. Meta cannot completely close the ecosystem because advertisers need to verify the effect across platforms, such as using Adjust to track the conversion of Meta ads, which naturally leaves a backdoor for AppLovin.

AppLovin asks its 600 e-commerce customers to invest $600,000/month ($4.3 billion budget) in Meta first. In essence, it uses Meta's ecosystem to verify which user groups have high conversion rates and then spends $1 billion (close to one quarter of total 2024 revenue) on the AppLovin platform to amplify the effect. Because AppLovin's advertising fee is low, advertisers can achieve the same effect on Meta by investing less money in AppLovin. This is indeed a smart and feasible strategy.

In summary:

• Meta’s Advantage：Accurately locate users, suitable for high-budget verification of user portraits, and stably earning $1.5 for every $1 spent.
• AppLovin’s Advantage：Low-cost traffic expands coverage, using Meta's data to find similar users among a large number of users, amplifying the effect, and earning $1.5 for every $1 spent.
• Mentality of Advertisers： First use Meta to test accuracy (spend a lot of money), then use AppLovin to test wider (spend a little money), and the two platforms complement each other.

Financials

Total Revenue & Margins

The total revenue of AppLovin in 2024 is $4.709 billion, a significant YoY increase of 43%. The profit margin also improves, with Gross Magin increasing from 67.74% in 2023 to 75.22%, mainly due to the technological breakthrough of Axon 2.0, the development of the e-commerce market, the increase in the proportion of Advertising, and cost optimization. The explosion of Advertising is the direct cause of the improvement in total revenue and margin, and the role of Apps is decreasing or even negative contribution. The second half of 2023 is a watershed, and Advertising replaces Apps as the core.

Source: Company Financials, TradingKey

Segment Breakdown

Advertising Revenue refers to the revenue earned through its advertising technology platform (formerly known as "Software Platform", renamed "Advertising" from 2024). This part of the revenue mainly comes from helping advertisers (such as game developers, e-commerce brands) to place advertisements, and monetizing their advertising space for app developers (such as games or tool applications) through real-time bidding and technical optimization. The full-year advertising revenue in 2024 is $3.224 billion (about 68% of total revenue), a YoY increase of 75%. This part is the fastest growing business of AppLovin, which is $1.842 billion in 2023, accounting for only 56%. Additionally, The Adjusted EBITDA for Advertising is much higher than Apps, which is more profitable. AppLovin's shift of strategic focus to advertising technology is also better for the company's development.

Apps Revenue refers to the revenue earned through its own applications (mainly mobile games). This part comes from game studios acquired or invested (such as Machine Zone, Lion Studios), including in-app purchases and in-app advertising. In 2024, the app revenue for the whole year was $148.5 million (about 32% of the total revenue), a YoY increase of 3%. This part of the growth is slow or even declining. In February 2025, AppLovin announced that it would divest its app business for $900 million to focus on advertising.

Source: Company Financials, TradingKey

Monthly Active Payers

Monthly Active Payers (MAP) is an indicator that is directly related to the Apps business. The average MAP for the whole year of 2024 is 1.6 million, a decrease of 11% from 2023, while the Average Revenue Per Monthly Active Payers (ARPMAP) has increased significantly by 10%. AppLovin has adjusted its in-game monetization strategy. By increasing the price of in-app purchases, launching high-end gift packages or limited-time activities, AppLovin allows a small number of high-paying users to spend more money, reduces low-paying small users, and focuses on high-value whales. The Apps business no longer attracts new users on a large scale but optimizes the existing user base. At the same time, the strategic focus has shifted to the advertising business.

Source: Company Financials, TradingKey

Valuation

We use EV/EBITDA to evaluate high-growth companies like AppLovin because it can fully reflect the company's high-growth operating efficiency, excluding the impact of debt, cash and non-recurring income, and is more stable than P/E. AppLovin's Advertising business has a high profit margin, and EV/EBITDA can better capture its core value, especially in economic fluctuations. AppLovin's 46.1x EV/EBITDA is much higher than the average (29.4x), mainly due to its high growth and high margin, reflecting the market's optimistic expectations for its AI and e-commerce pilots, but the fact is that AppLovin's rapid revenue growth in recent years is partly dependent on its investment in AI technology, and partly benefiting from the natural growth of the global digital advertising market.

At present, concerns about a recession in the United States are growing, which may pose a significant risk to AppLovin's revenue growth. In the economic downturn cycle, governments often cut public spending, corporate revenue is under pressure, and advertising spending is reduced accordingly. In order to control costs, companies will reduce bidding advertising budgets, especially high-risk programmatic advertising, which directly affects AppLovin's core business. Once the economic growth rate slows down, AppLovin will face not only a slowdown in revenue, but also market doubts about the sustainability of its growth. The above-expectation performance in 2024 is driven by the dual forces of AI and the economy. If the economic pillar collapses and the natural growth of the market disappears, investor confidence will inevitably be affected, and then EV/EBITDA may have the risk to return to 30-40x.

Source: Company Financials, TradingKey