The AI-powered recommendation engine behind AppLovin's performance-based advertising solutions
AXON AI, our AI-powered advertising engine powers AppDiscovery, the cornerstone of our advertising solutions. It enables the matching of advertiser demand with publisher supply through auctions at vast scale and at microsecond-level speeds, helping advertisers maximize the return on their advertising spend.
How does AXON AI work?
AppLovin advertisers use AppDiscovery (and our e-commerce solution, which remains in beta) to automate, optimize, and manage their marketing efforts.
Advertisers set specific return goals for their advertising campaigns.
AXON AI's predictive algorithms then evaluate potential impressions with the goals of meeting advertisers' return on advertising spend targets and other marketing objectives and finding the users most likely to download or otherwise engage with their product(s).
AppLovin bids on impressions based on AXON AI's determination of the value of the impression to the particular advertiser.
AppLovin earns revenues based on the value we drive for advertisers as measured by third-party attribution, not simply a fixed price per impression or per action (click or installation). Our performance-based advertising model therefore demands that ads served by us lead to genuine, high-intent engagement, ensuring our campaigns deliver meaningful, measurable results for advertisers. AXON AI's algorithms power these outcomes by making predictions about the users most likely to download or otherwise engage with their product(s) and submitting bids for impressions served to those users.
Additional Details
Complex AI models, like AXON AI, train and optimize on a continual basis using an assortment of data. Our models use device and network information and other information we have collected, or our partners share, about users' engagement with applications, ads, and other digital properties, as described in our Privacy Policy and How AppLovin Shows You Ads page. In addition, our models use win or loss notifications from our in-app bidding and monetization solution known as MAX, which all bidders receive depending on whether the bidder won or lost a particular impression, as well as advertisers' audience data if they choose to share that data with us.
We do not collect or seek to elicit—and our models do not ingest—users' "sensitive personal information" (such as social security numbers, precise geolocation, racial and ethnic characteristics, religious or philosophical beliefs, or the contents of messages) or protected characteristics (such as age and gender). In addition, we do not collect or seek to elicit—and our models do not ingest—other bidders' bid or user data. We also explicitly prohibit partners from initializing our technologies in connection with children's data or otherwise providing us with children's data, and we prohibit apps exclusively designed for and/or exclusively directed to children on our platform.
For more information, please review our Privacy Policy and How AppLovin Shows You Ads page.
On an ongoing basis, AXON AI's models experience compounding improvements through self-learning, continuing data ingestion, and engineering oversight and enhancements. As more advertisers use our performance-based advertising solutions, the models improve through additional data and better insights (per the collection described above), which then further enhances the efficiency and effectiveness of our performance-based advertising solutions.