Apple Mail Privacy Protection Findings & Update

How Email Marketers Need to Adapt Their Measurement & Omnichannel Strategies

Background

Apple released its latest operating system update, iOS 15, in September 2021. This was a closely watched update as Apple introduced new email privacy features. Most notably, iOS 15 features “Mail Privacy Protection” (MPP) which hides user IP addresses and preloads email content remotely on the user’s device to prevent senders from seeing if the email was truly opened. This update has posed a challenge for email marketers by hindering their ability to clearly measure email opens, as a large percentage of ‘opens’ are instigated by the technology and not the user.

Prior to the release of the update, there was speculation about how Mail Privacy Protection would affect email marketers. CMI Media Group has been closely monitoring and analyzing email performance data to measure the impact of iOS 15 and provide updated findings and recommendations.

Even as iOS 16 looms, marketers are analyzing the effects of the privacy changes related to iOS 15. This POV reviews those changes, the impact, and what brands need to know.

Impact

While initially iOS 15 was not adopted as quickly as anticipated, Apple now reports that 63% of iPhone and 49% of iPad devices are now running iOS 15. After downloading the iOS update, or activating a new Apple device, users are prompted to enable Mail Privacy Protection upon opening the native Apple Mail app.

The privacy features can be enabled for the native Mail app, but do not appear to affect other emails apps, such as Gmail or Outlook. While Apple has not disclosed how many users have enabled this feature, based on internal HCP email performance data, CMI Media Group estimates that about 45-50% of users are impacted. Because the email privacy features of iOS 15 pre-load mail content before the email is opened by a user, an “auto-open” occurs which causes email open rates to appear inflated and makes it difficult to measure true opens amongst iOS 15 users.

In email performance data analyzed by CMI Media Group, we have observed that impacted users appear to have their browser or operating system values listed as “unknown.” In November and December 2021 as adoption of iOS 15 increased, we noticed a large increase in “unknown” agents in email open data, that was not prominent in our data before the iOS 15 release. Internal testing revealed that these “unknown” users are likely utilizing iOS 15. While not all opens by “unknown” users are false, looking at this data point does help us to estimate the potential scale of impact.

CMI Media Group has been meeting with the media supplier partner community to understand what email service providers (ESPs) they are using and how each media supplier and their ESP are managing inflation in email open rate data. Some ESPs have begun to introduce data filtering to account for false email opens, while many others, including the predominant Salesforce Marketing Cloud, have not yet implemented any monitoring or filtering functionality. CMI Media Group has requested that all media suppliers provide raw email performance data, including platform and browser type, without any filters (other than standard non-human traffic exclusions) for analysis to do a true comparison across vehicles.

Between Q1-Q3 2021 and Q4 2021-February 2022, CMI Media Group has observed open rate increases of 6-8 percentage points on average for 3rd party media suppliers and 3 percentage points for our proprietary owned email service, iNBOX. There are extremes on both sides where some suppliers are seeing a decrease in open rate, which may be a function of their Email Service Provider (ESP) and how they are handling the iOS 15 changes, while others are seeing >+100% change. The range of impact varies greatly from -54% to +305%.

Average Email Open Rates	Q1 ’21 – Q3 ‘21	Q4 ’21 – Feb ‘22
3rd Party Media Suppliers	8% – 9%	14% – 17%
iNBOX Owned Emails	9.3%	11.9%

Recommendation

MEASUREMENT
From a measurement perspective, CMI Media Group recommends the following approaches to address iOS 15’s impact on email performance.

Assessing browser and operating type data to identify “unknown” users is currently the best approach to estimate the number of users who may be impacted by MPP changes. Email recipients with non iOS and known browser and operating system types will provide an analog for email open performance. Additionally, we recommend looking beyond email opens, and measuring other data points which have not been impacted by the iOS 15 update, for a clearer picture of email engagement:

• Monitor email open rate for “unknown browser or operating system” users vs. other user types. Create different sets of benchmarks for each user group, expecting “unknown browser or operating system” users to report higher open rates.
• Prioritize other email performance metrics and shift away from using email opens as the only success metric. Consider other metrics such as delivery rate, opt-out rate, and click rate to measure email engagement and performance. For iNBOX emails, these rates have remained consistent despite the change in open rates during the iOS 15 update. Deliverability averages 95%, opt-out rates are less than 0.5% on average, and click rates average 1-2%. The Mail Privacy Protection changes have not impacted on our ability to reach our audience.
• Prioritize engagement drivers within emails, such as links to videos or helpful resources, where engagement metrics like views and downloads can be accurately tracked.
• Tag brand websites with 1st party pixels which allow marketers to measure on-site engagement for users that click through to the website and to tie back their activity to previous website visits.
• Focus on building long-term relationships with customers, rather than the short-term impact of single emails. Emphasize the importance of quantifying the long-term impact of email marketing, by measuring lagging indicators like Rx behavior.
• Ensure transparency when comparing email data across suppliers and systems. When collecting email performance data from internal systems or 3rd party media suppliers, ask if the supplier or their ESP is applying any filtering or modeling to the email performance data. Some ESPs have begun to roll-out solutions for estimating true email opens, and some suppliers are creating models on the back end; both of these will result in an incongruent comparison of data.

EMAIL & OMNICHANNEL CAMPAIGN RECOMMENDATIONS
The iOS update further emphasizes the need to have a cohesive marketing strategy across multiple channels. An omnichannel marketing strategy, where the coordination of multiple channels can be executed, can help mitigate the impact of changes to any single channel.

For clients who are engaging in omnichannel marketing programs, such as those utilizing CMI Media Group’s proprietary PROACT™ platform, which can monitor and react to email opens (among other NPP channels) to trigger next best actions in the campaign, may want to modify campaign business rules now that email open data is less reliable.

• Consider modifying the engagement metric from opens to more measurable actions in omnichannel or triggered marketing campaigns which monitor tactic engagement to determine next best actions. For example, engagements can be measured as a click, read time, video view, contact request, or other call to actions.
• When launching echo email resends to non-engagers of an initial email, consider sending echo emails to “non-engagers” rather than “non-openers.” This ensures proper reach and safeguards recipients who have engaged with the email from receiving the email again. It also provides “auto-openers” another opportunity to engage, if they did not truly open the initial send.
  • However, changing the engagement metric from opens to clicks or other actions, does risk deploying a duplicate echo-email to iOS 15 users who truly did open an email but did not perform the desired engagement action. To minimize potential negative impact from duplicate emails, we recommend utilizing a different type of tactic, different email sender, and/or different email subject lines for any echo email deployments.
• Incorporate non-email media tactics into triggered or omnichannel campaigns. Many 3rd party media suppliers offer alternatives to email, such as onsite or app-based alerts, that can deliver a similar type of message as email, but in a 3rd party environment which is more accurately measured.
• When sequencing tactics in a triggered marketing campaign, prioritize non-email tactics to be deployed prior to emails in so next best action decisions can be made based on more accurate engagement metrics earlier in campaigns.

Conclusion

The need to evolve how marketers strategically target, track, and measure programs is just beginning as the movement by Apple to ensure privacy of its users, and obscure data from marketers, is only the beginning. The ability to track and target at an individual level, especially without first party data, will become increasingly difficult as privacy legislation evolves at the state and federal levels. Creativity will continue to be married with data and technology to find solutions around these roadblocks.