Meta has shared a new overview of some of the latest AI-powered updates that it’s made to its app advertising options, and how they’re driving better response for various developers.
Meta says that its improving models are driving better performance in all elements of app promotion aims, while these latest advances also showcase how its AI tools, in general, are driving better ad performance, in new ways.
For example, for app advertisers aiming to drive in-app purchases, Meta’s AI-powered optimization approach offers two aims: Driving conversion volume, or focusing on the most valuable customers.
For most app advertisers, aiming for download volume would be the logical choice, but Meta says that its improving AI targeting has now made value optimization even more effective.
As explained by Meta:
“Throughout 2025, we have continued to enhance our models to make value optimization, also known as ‘maximize the value of conversions,’ even more performant. Today, advertisers focusing on optimizing the value of conversions in app campaigns see a 29% higher return on ad spend (ROAS) compared to those optimizing for volume of conversions, making this the most performant Meta solution for App advertisers optimizing for ROAS.”
This is an example of how an AI system’s capacity to process billions of data points can help to pinpoint more valuable ad outcomes, based on a broader scope of data points that human assessment would not be able to measure.
In other words, Meta’s system is now able to input many more data points on each user, and cross-match them against other behaviors to find similar audience matches, which is driving more valuable outcomes over traditional focus aims.
On another front, Meta says that it’s implemented a range of AI enhancements, in collaboration with its mobile measurement partners (MMPs) to help advertisers optimize their ads to deliver more of the conversions that advertisers value in their third-party analytics tools.
“For example, advertisers may have distinct definitions for specific categories or events within their MMP reporting, such as how long after a first app install a user can be considered as a ‘new user’ if they reinstall the app, or how long a user must be inactive in the app before they can be attributed to a reengagement campaign. By gaining insight into these varying timeframes, we can more effectively pinpoint and reach the types of people that each advertiser deems most valuable.”
Using this expanded insight to help build its AI model, Meta says that it’s now updated the window that advertisers use to determine what counts as a new user that should be targeted with app promotion campaigns.
“Advertisers utilizing AppsFlyer can leverage a flexible reattribution window spanning from 0 to 180 days reflecting how it is set in AppsFlyer. Conversely, Meta will apply a 180-day exclusion window for its app promotion campaigns for advertisers working Adjust and Singular, which employ a lifetime new user window.”
Meta says that extending the exclusion window has seen Mobile App Promotion campaigns targeting new users drive an average of 20% less acquired users that did not meet the advertiser’s categorization of a new user.
The option is now available through Meta’s various app promotion partners.
