Search specific term/phrase surrounded by double quotes. e.g. “deep linking”
Exclude records that contain a specific term prefixed with a minus. e.g. Android -Firebase

What happened to Predictive Modeling? In which situations is it still an advantage?

Predictive Modeling very much still exists, and will continue to power a number of critical use cases for our clients. These can be thought of as use cases that fall into two different buckets: paid ads and organic/owned links.

1. Paid Ads. When users have agreed to tracking via ATT in both the publisher and advertiser apps, the vast majority of device-level attribution will be done via IDFA. However, there will still be edge cases where this is not possible (such as the IDFA not populating in the tracking link). In these cases, Branch will be able to use Predictive Modeling to accurately attribute that user to the proper ad.

Additionally, since the IDFA is never accessible on mobile web, Predictive modeling will power all web-to-app ad attribution (so long as the user opts in via ATT in the destination app).

2. Organic Channels. Since Apple’s new guidelines apply to paid advertising, organic channels can still continue to function as they do today. Whether it be app-to-app links, links on a client’s site, links in an email, or non-paid links on another site, Predictive Modeling will power best-in-class attribution.


Additional Resources