This week, mobile measurement firm Branch Metrics released data from their network related to the adoption of variants of iOS 14.5. Unsurprisingly to everyone operating in mobile, the adoption rate curves of iOS 14.5 and 14.5.1 look very different from previous releases:
The curves for iOS 14.5 variants above are shaped differently than the adoption rate curves for other releases, especially given that 14.5 was a momentous release that included a new feature (App Tracking Transparency, or ATT) that was originally designed for inclusion in 14.0. The inflection present in most of the curves above represents the point at which Apple triggered a system notification for users that a new version of iOS exists: this is when most users instantiate an update.
iOS 14.6 was released to the public on Monday. But why did Apple slow-roll the adoption of variants of iOS 14.5, which unveiled (arguably) the feature that most impacts the mobile app economy since the launch of the App Store? (update: after this article was published, Branch released iOS 14.6 adoption data)
I have a few theories:
- Apple needed time to address unexpected behavior with the “Allow Apps to Request to Track” setting. Many users discovered upon updating to iOS 14.5 that a new setting, “Allow Apps to Request to Track,” which controls whether apps can expose the ATT prompt to users, was defaulted to off and disabled. Keep in mind that the IDFA is only made accessible to an app when the user opts into ATT after seeing the prompt (in other words: the IDFA is inaccessible by default). Apple explained that certain classes of devices (those used by children and educational organizations) should have this option disabled, but this limitation applied to a larger group of devices than that in iOS 14.5. If this unexpected behavior needed to be fixed, it makes sense that Apple wouldn’t stimulate the adoption of iOS 14.5 and variants;
- ATT’s effect on the mobile ecosystem will be expansive and unpredictable, and Apple didn’t want to shock the ecosystem. CPMs for opted-out iOS 14.5 users are, as expected, much lower — in the case of one large publisher, 50% lower — than CPMs for opted-in users. If iOS 14.5 had been adopted at the same pace as previous prominent releases, it could have caused havoc for advertisers and publishers, many of which were ill-prepared for the changes being implemented with ATT (in part because many ad platforms hadn’t yet completed their SKAdNetwork integrations when iOS 14.5 was released). It absolutely is logical to roll out a high-impact feature at a measured pace to give developers time to adapt to it, and to dampen its reverberations across the developer ecosystem.
But a corollary to that second point should acknowledge that a slow rollout obfuscates the impact that ATT will ultimately have on the mobile ecosystem. Almost every company that ATT touches — advertisers, publishers, ads platforms, and even Apple — is incentivized to downplay the impact that ATT will have on their business. Advertisers want to promote the idea that they can maintain advertising efficiency without access to device-specific behavioral profiles. Publishers want the same. Ads platforms want advertisers and public markets investors to believe that they didn’t disproportionately utilize device-specific behavioral profiles for targeting in the first place, or that contextual or on-platform data can be used to the same effect.
And Apple wants to promulgate the notion, especially as it expands its own ad network, that ATT really only provides consumer benefit by eliminating “tracking,” and that no countervailing force is being created with respect to consumer choice or detriment to developers. If ATT is rolled out slowly, its impact on the ecosystem will be revealed equally slowly and won’t make headlines that call Apple’s motivations into question. The frog that’s being boiled doesn’t scream for help.
Edit: the article originally stated that iOS 14.6 is still in public beta, but it was released to the public on Monday, May 24th
Photo by LOGAN WEAVER on Unsplash
Recent