Apple launched their new privacy measure, Apple Mail Privacy Protection, in mid-September 2021. Along with that update came a set of new privacy measures, specifically Apple Mail Privacy Protection which discontinued opens and IP Address tracking. Because of this, the majority of emails sent to Apple devices will appear as opened and inflate open rates if reports are not adjusted.
This article will cover how Persado will measure your KPIs for Experiments in light of this change. We have found that this Apple iOS change will not impact our ability to accurately review results.
Background: the Impact to Customer-centric Marketers
Apple’s version 14 of its iOS operating system for iPhones and iPads, launched in July of 2021, included a number of new features that make it harder for third parties to gather user data. Most of the conversation has focused on the loss of third-party cookies — a major shift for sure, but not the whole story.
Beyond cookies, Apple is also blocking the work-arounds that app owners have developed to supplement (or compensate for the loss of) cookie data. Known as App Tracking Transparency, or ATT, Apple’s iOS14 additions require users to opt in to allowing third-party apps to track them. When a user gives permission, ATT requires the app to document and share (with the user) what it tracked.
The bottom line is that with ATT, many of the techniques marketers have used to gather user-identifiable data from an Apple device, including fingerprint identifiers, IP addresses, location data, or device characteristics, are off limits unless the user expressly allows them.
The Impact to Persado’s KPI Analysis
In a nutshell, Apple Mail’s Privacy Protection may impact a small proportion of Persado customers who Experiment on very small audiences where opens is the only metric for which we can reach significance. Although we would lose this minor ability, it is offset by Persado’s Predictive Content capability which does not require us to reach significance. When opens are not the only metric, we can use clicks as our primary metric.
84% of the time, Persado reaches click significance in our subject line Explorations, nullifying the need for opens.
Our Data Science team also conducted a comprehensive analysis to understand if open rate patterns for Apple users match those seen across all users. They discovered that Apple users respond to content in a highly similar way to non-Apple users, which means that excluding Apple users when calculating open rate won’t impact our ability to run an analysis on how individual parts of a message, or “genes,” are performing. Check out their analysis below for more details.
Apple vs. Non-Apple User Analysis
Our Data Science team ran an analysis on 184 Explorations with information on what Operating System was used for the first open and first click to see if open rate patterns for Apple users matched those seen across all users. As open rates skew upward, they found that Persado’s ability to track incremental changes in open rate will remain stable.
In the calculation above, our Data Science team excluded Apple Mail opens from their calculations, so incremental opens are inevitably reduced. However, notice how the open rate for Variant A (7.5%) is still higher than the control Variant open rate (5%) when excluding Apple Mail opens. In summary, Persado is still able to conduct its usual gene analysis on the opens metric, and the winning content will be the same.
Key Takeaways
Excluding Apple users will not impact results. Apple users respond to subject line content in the same way as non-Apple users. Even if Apple users are excluded from our results, Persado’s analysis generates nearly the same output.
Predictive Content can address outlier Experiments with very small audiences where opens is the only metric for which we can reach significance. Because we don’t need to reach statistical significance with Predictive Content, we can use this methodology instead of Experiments when needed. Data Science and the Brand Content Strategy team can help guide you in these instances.