Skip to content
Ideas Measurement

Apple's iOS 14 release and what's next for in-app measurement

Eric Kirtcheff

Apple doubled down on Privacy at its annual WWDC by stating “Privacy is a fundamental human right” to open its privacy product roadmap. They highlighted four core privacy principles of Data minimization, On-Device Intelligence, Security Protections, and Transparency & Control.  All of which could limit marketing data shared from mobile devices. This product roadmap upholds the GDPR & CCPA laws by asking user permission to be tracked and providing transparency of what data will be linked. Here is a deeper dive on some of the announcements from the documentation:

In-app tracking

In-app tracking will move to opt-in just like location based data on iOS today. More significant is that it’s one-time only. This means it’s important to communicate clearly what is absolutely necessary upfront to maximize consumers giving their consent. Marketers should consider why the user would choose to opt-in and not just assume they will or won’t.

Developers will need to focus on the value proposition to use the app.  Opting in gives free access to using the app or the user can pay for an ad free experience. It may be easier to digest than web counterparts, where you need to accept cookies each time to read one article. 

Identifier for Vendor (IDFV)

One of the changes under the Transparency & Control principle is the loss of IDFA (Identifier for Advertisers) which was assigned at the user device level and the move to Identifier For Vendor (IDFV) released in a 2012 update, which is a device ID unique to each vendor. IDFV is similar to IDFA but the ID is created at the Vendor level, so Google owned Youtube would have a unique IDFV from Google Search App. Monetization using attribution might be in jeopardy, as publishers see a 50% drop in revenue without IDs, but advertisers could use IDFA for First Party Analytics. IDFV will allow individual apps to track using identifiers at the vendor level within their apps/properties. Large Scale apps should be able to track within their ecosystem, small scale apps looking for third party ad network support will lose out on measurement. Perhaps measurement will revert to daily active users as more of a branding play.

SKAdNetwork

SKAdNetwork is Apple’s version of Chrome’s Privacy Sandbox or Safari’s Storage Access API (ITP). You can essentially think of this as a combination of APIs that allow marketers to track app installs on iOS. However, a big unknown question here is how they will handle users who interact with an ad but do not download the app. They do not call out if advertisers will be given tools to run engagement campaigns across third-party apps without IDFA, report, or create audiences from these exposed users who haven’t installed.

Measurement Impact

Early estimates from Essence’s Ad Tech Team indicate that these changes could reduce measurement in-app on iOS by as much as 80%. We know this from the adoption and impact of prior ITP releases. For advertisers that rely on in-app measurement across iOS devices, Essence’s Data, Tech and Analytics teams can help ensure plans are in place to maintain the highest levels of measurement.

We can also conclude from these changes that location data will decline by about 10-20%. However for the portion that remains, we think the value of the location data will increase (eg. The Weather Channel). This could make for some interesting new strongholds in the mobile ecosystem for location. These changes are happening as we see the rise of 5G, creating an ultra-dense network which implies more access points that could be trackable. Perhaps location data providers will spring up again through this new layer of connectivity.

Lastly, we predict a rise in companies doing on-device measurement with some sort of incentive in exchange for the data/metrics.

What now?

We suspect many advertisers will be adjusting their reach potential in Q4, where we could see a loss of about 20% of reach just related to how in-app device IDs are measured. One alternative we have had success with is the Google Ad Manager Product Non Personalized Ads (NPA). This product uses models to determine frequency using contextual or coarse level geo (city-level) means to enable tracking. It’s a good tool to have handy in areas where measurement is not always possible, though it’s important to note that Audience or User List Targeting is disallowed.

All of these changes are going to be an adjustment for any measurement reporting. Year over Year comparisons of data will not be the same. In some cases, viable alternatives will be tested and perfected. In others, metrics will have to be analyzed at higher grains. At the end of the day, it will still be possible to see increased sales, traffic or stream requests for content, but the ability to tie it back to an IDFA or device ID will be lost.

Dan Elddine contributed to the writing of this article.