Why Apple’s Email Privacy Announcement Should Make You Rethink Your Email Strategy

Last week, Apple announced a series of new privacy-preserving features, and for email marketers, there was monumental news:

This fall, senders will lose open, IP, and location tracking for all Apple Mail users - regardless of the email service provider the app is connected to, including enterprise email.

How big is this? Well, our friends at Litmus estimate Apple Mail accounts for about 46% of email opens globally. 

With potentially dramatic implications, this can—and should—be the catalyst for us to take a different approach to effective engagement marketing: 

Generally, we need to send fewer emails that are highly relevant, highly personalized, and optimized for click-through and business impact — not opens.

While there are still many questions about how this will work technically, the coming changes have been emerging over the past few years, especially related to proxy scanners. Our AI algorithms have identified a growing number of simultaneous, automated "opens" in our customers' contact databases, indicating this is not limited to Apple. 

Let's discuss the Implications — first, what's changing with Apple Mail Privacy?

Again, to summarize the changes, senders will lose the majority of tracking for any emails opened in Apple Mail. Simon Harper has an in-depth community-sourced public doc going that outlines what's changing for Apple users who opt-in, and the most significant changes are:

• Tracking pixels will get fired upon email delivery, along with all other remote images 
• IP addresses will be blocked
• Location will be blocked
• Email forwards: you guessed it, blocked
• Actual email addresses may be obfuscated behind an Apple proxy email address
• Header information will be limited, 
• Client and device type will be obfuscated

Yes, this only affects contacts using Apple Mail (likely up to half of your opens). Still, you should expect to see more of this in the future from other consumer email providers like Outlook, Yahoo Mail, Google, etc, and all enterprise email systems (ProofPoint, Barracuda, etc). 

How does this affect your email marketing?

Given that open signals and their timestamps will be lost, you will see three main changes to your email marketing:

1. Open rates cannot be an anchor email measurement. 

Reality check: this has been happening for years with the big commercial email providers and increasingly in enterprise email systems. Engagement metrics that rely solely on those open signals or timestamps will be increasingly meaningless.

2. Your tools, sequences, and workflows may break significantly:

Carefully constructed open activity triggers of various kinds will get fooled and route any proxied "opened-engaged" contacts immediately. 

If you have multi-step campaigns, lists, or logic that depend on these signals, you'll need to rethink your approach. The usual open time-based send time optimization (STO) methods commonly used in the industry will stop making any difference. The same is true for fatigue analysis and throttling, as it's widely done today.

3. Open rate A/B testing can no longer be the sum of your testing strategy: 

As Brian Sisolak notes, open rate-based A/B testing will become less effective overall and useless for any known contact sitting behind a proxy server. Open rate-based list hygiene gets complicated fast and may result in contacts staying in email marketing flows for too long.

That doesn't mean you can't do testing (see below). But your UOR-based A/B subject line test will be polluted with phantom opens and skewed results.

4. Everyone needs to up their deliverability game.

You still have to get all the deliverability basics right - DKIM, DMARC, SPF, sender reputation, active blocklist management, etc. If you don't have someone on your team accountable for this, get someone or outsource it.

Starting in 21Q4, deliverability will take on additional dimensions as Apple, and other email service providers adopt additional AI-driven quarantine and classification. These changes will introduce more uncertainty into contact activity streams — both false positives (illegitimate open activity) and false negatives (silent quarantine returned as delivered).

Recommended changes to your email marketing practices

For many email marketers, this may seem like one of the four horsemen of the apocalypse. Still, if you care about your work and want it to drive actual results and business impact, this can be the catalyst you need to drive organizational change. 

Expect similar shifts with other email service providers, and the most aggressive ones to adopt will be enterprise email systems (many already have). Expect to operate in a future where personal data has been obfuscated or is absent. 

If tracking is blocked until the recipient hits your first-party content, we can look to performance marketing best practices on how to engage at scale and with efficiency:

1. Earn your right to engage through real personalization, high-quality, relevant content, frequency control, and timing. 

High-level: think about pull over push. If email recipients are in the driver's seat and response data is nil, email marketing is no longer about pushing as much out as possible.

This likely means we will all need to send way fewer, highly personalized emails — and Apple's change gives you the business case to push back on stakeholders who want you to batch and blast over and over again. 

2. If you're relying on open rates as your primary engagement performance metric, stop.

A better approach - depending on what you're trying to measure - might be unique click-through rate or click rate or (better yet) aggregated multi-channel activity against time.

Be aware that there is also already phantom click behavior from some proxies doing full email scans, which we expect to worsen over time. 

3. Design your nurture strategy around click-throughs, web visits, form activity, purchases, and other demonstrable, first-party interest data.

Think in terms of building behavioral "signal resilience" in a world where you'll need more kinds of data with no 3rd party cookies and limited email tracking.

4. Highly tailored and targeted campaigns matter more than ever and must adapt and evolve.

Find opportunities to create highly personalized emails based on a combination of activity, demographics, and other 1st party data.

5. List hygiene will need to become more proactive based on a deeper analysis of contact and proxy server behavior.

Don't just rely on opens to tell you this (but this was already true).

We know that while many marketers may pay these tactics lip service, for many, they appear out of reach due to headcount or data limitations. 

We can't solve this problem by adding more headcount to build more static sequences or more batch-and-blasts. 

The only healthy long-term solution is to leverage analytics, AI, and real automation at scale — just like we do with every other performance marketing channel at the top of the funnel. 

How will this impact Motiva AI users?

Motiva users will see little or no change and have access to our full range of machine learning and analytics capabilities.  We've also been pre-emptively working on a new product that is helping our customers identify and actively manage this growing trend of ESPs scanning mail and sending ghost engagement signals back to ESPs. 

A few key callouts:

1. Send time optimization is still relevant, but your current STO system will become useless if it relies on open rates to function. Because of the way Motiva's Send Time AI was built, customers will continue to see engagement rates improve even with polluted open data. Motiva incorporates a variety of signals beyond individual opens.

2. Message testing is more critical than ever. Many of the hot takes we've seen tell email marketers to give up testing things like subject lines, sender personas, or email preheaders. 

This is nonsense. If anything, this means testing messages is more critical to marketing success than ever, but you'll need to pick optimization goals other than Open Rate to do this. Motiva customers will still have full automated A/B/n and multivariate testing options available.

3. Targeting always matters, but take ongoing segmentation, persona development, and targeting more seriously than ever. If you are in a low data quality environment, you have to assume your first cut at a segment should be improved over time. Use tools like Motiva to analyze those definitions and surface improvements.

4. Frequency management and fatigue management become even more essential to protect deliverability. It's critical to adapt to your contacts' preferences, and you have to take an approach that isn't completely reliant on open activity, which we've done at Motiva.

What's next?

While this announcement ushers in broad changes to how we think about, measure, and deliver emails to our customers, we believe it will be a net positive for consumers and email marketers. We'll continue to monitor this and related privacy news and offer guidance alongside our industry-leading tools. Sign up below if you'd like to stay in touch on this and related topics.

As email marketers, we can't afford to keep going with business as usual if we want to maintain email as an effective and high-impact channel.

Use this opportunity to move your email marketing into 2021—by supercharging your team with the best of AI, so they have the time to build strategic, personalized, and dynamic content that drives actual business results. 

Being customer-centric always wins, and it's time to ensure our email marketing delivers on that promise.