Meta has announced significant updates to its AI advertising optimisation system for mobile app and game advertisers.
The main innovation is improved conversion value optimisation (Value Optimisation / Maximize the value of conversions), which focuses not on the number of installations, but on attracting users who perform the most profitable actions, such as purchases and subscriptions.
According to Meta, the new algorithms deliver a 29% higher ROAS compared to campaigns that target conversions. Early materials and reviews for 2025 mentioned an increase of around 12%, which indicates rapid progress in Meta’s machine learning.
The company has also improved closer integration with mobile analytics platforms (MMPs) such as Adjust, AppsFlyer, Kochava, Branch, and others. This has helped reduce discrepancies in reports and more accurately account for the criteria advertisers use to define ‘new users.’ As a result, advertising campaigns have become more accurate and effective, with less budget waste.
These updates are particularly effective for the gaming industry, where it is important not only to attract players, but also to retain them and monetise them.
Meta’s new features are part of the company’s broader strategy to automate advertising (Advantage+, Value Rules) and strengthen competition with Google and TikTok, which are also developing AI tools and solutions for confidentially measuring advertising effectiveness.
This way Meta is focusing on audience quality and real user value rather than the number of installs, helping developers to more accurately manage the profitability of their apps, especially if they promote paid content.
- Read more: Meta’s blog
