A 6-month retrospect - How have we dealt with Apple’s Mail Privacy protection changes?
The journey to find metrics in a metric-less world
Last September, Apple completely changed the email marketing landscape by shifting the way their client handles audience privacy.
The proposed changes in Apple Mail sent waves across the marketing world. Between cries of “email marketing is doomed!” and “open rates are dead!” there was a bit of panic and insecurity as everyone tried to scramble in figuring out new strategies to replace their age-old, time-proven tactics.
Perhaps with good reason. This wasn’t a small change. Apple Mail is used by a considerable portion of the global audience and if the rumours are to be believed, these measures might soon be adopted by the likes of Google.
So we paid attention, observed, and adapted – and now, six months down the line, we can see what has worked, and what hasn’t.
Let’s take a moment to recap.
What changes were brought forward by Apple, and what effects did it have on traditional email marketing?
In essence, from September onwards, if a user opted-in to Apple’s new privacy protection, the EMS client would lose visibility of their opens, and as such, it would register as a false positive.
In just one update, we lost track of a key metric across one of the biggest email service providers in the world.
And, I suppose, it would be one thing if this happened across the entire spectrum of email marketing. But for better or worse, it only affected a certain section of the audience.
Let’s see what areas got affected.
- A/B testing of subject lines – Anything that relies on metrics from open rates is now under scrutiny, and testing subject lines has now become a much more tricky endeavor.
- Click-to-open-rates – CtOR was the innocent bystander caught in the shoot-out. A relevant and key metric that we used in our reports as much as we did open rates, and unfortunately, was negatively impacted by the changes.
- Targeting/automations based on open dates/open-date frequency – Obviously anything that has to do with how often, or not, a user opens a campaign became difficult to track. Which had a knock-on effect in that pruning an audience base to ensure deliverability became much more difficult.
- Send Time Optimization (STO) – Seeing as we’re now unable to correctly record when/if an email was opened, optimizing email sends based on open times was also affected.
- Identifying device type/location – With such data now being considered private, and not offered by Apple, we lost the ability to accurately define the location and device type of our audience.
- Cooldown timers based on opens – Unfortunately, this feature was also affected and would show the time when the software scanned the email rather than the time it was opened by the user.
- Non-opener resends – A useful tool for any marketer wishing to promote a specific campaign, which was also negatively impacted by the change.
- Long-form emails – A change that was quite impactful, yet unobserved at the time. Long-form campaigns that rely on visibility rather than clicks will not be used as much now, due to the inability to measure any relevant metrics.
Did we witness the death of Open Rates? And does it even matter?
The answer, much to everyone’s disappointment is – it depends.
Getting information from relevant metrics is still possible, but finding them has become much more difficult. Using open rates for information is now akin to trying to catch a slippery eel, but it’s still very much doable – albeit, you’d have to use a great number of workarounds.
Moreover, seeing that Apple possibly started a trend that will possibly soon be echoed across the industry, the more important question to ask is – is the change big enough for us to care?
And frankly, it’s worth considering. Having said that, this is a relatively new concept, and we’ve all been trying to find ways to adapt until the dust settles – here are a few ideas you can adopt in the meantime.
Measures you can adopt
1. Stop chasing windmills and start investigating patterns
I think the mistake marketers often make is putting the cart before the horse.
Goodhart’s Law is often in effect here – “When a measure becomes a target, it ceases to be a good measure.”
Put down the pitchforks for a moment. Quite often we’re so invested in our metrics because they showcase our relative worth as professionals. We want to be able to show how great of a job we’re doing, or how good our brand is, and we view these numbers as the holy grail. But they’re in fact, a tool, meant to help us reach our goals.
So, maybe a measure of rethinking is in order.
Obviously, metrics are important. They’re a tool to both determine our success and shape our strategy. But maybe we should adapt the way that we process them.
While, overall, the numbers showcasing our exact open rates aren’t reflective perhaps of reality, the shift in these numbers is still accurate. Instead of defining success or failure in terms of a total percentage, we should shift our reasoning around relative growth or decline on a month-to-month, and year-to-year basis. We should still be able to determine success and failure and adapt accordingly based on shifts in OR and CtOR. Don’t rely on the numbers up front, but investigate the patterns behind them and you will find your answers there.
2. Change your purging process
Keeping a clean and healthy audience base is a key part of the email marketing process. But what do you do when the metric that judges audience engagement goes away?
Obviously, if an email marketing contact hasn’t opened any campaign you’ve sent in 4-6 months, they’re not an engaged party. As per best practice guidelines, you remove them from your audience list to keep your metrics alive and relevant.
So, how do we adapt now that we can’t use open rates as a qualifier? It’s a tough choice, but we should be moving on to clicks.
Yes, it’s a much more costly solution, but if implemented properly, it shouldn’t cause any spillage. True, it demands a bit more from the audience, but if you also develop a proper and detailed sunset campaign with multiple potential saving points for an audience member to take advantage of, you should have nothing to fear.
3. Avoid long-form campaigns
Long-form email content has its advantages – you can increase audience engagement by ensuring they can interact with your content on their terms, without having to navigate through to external sources.
Unfortunately, while previously we could somewhat measure the relative success of such campaigns through open rates, that window of opportunity is fast closing. With no discernible way at the present of determining the effectiveness of long-form campaigns that don’t rely on clicks, the best course of action is to move away from them altogether.
4. Re-evaluate your metrics across the board
With open rates and click-to-open rates now being less useful as a metric, how can you continue measuring success? It’s time to broaden your scope.
We got bounce rates, clicks, conversions, unsubscribes, spam complaints, lifetime value, and ROI – all useful data that can be used to gain valuable insight. Just because we’ve been relying on open rates from the beginning, does not mean that we’re without options.
Finally, try to encourage your userbase to reply to your campaigns. Engage in surveys. Use that engagement as a further metric. Think outside the box. The more data you can pull, the better work you’ll produce.
5. Start using preference centers
Speaking of outside-the-box thinking, consider the following. Launch a detailed automation campaign that gets your users to supply you with their information using a preference center. Ask for information that pertains to their email client of choice, and then use that data to isolate the portion of your audience that uses Apple Mail.
By further investing in separating your audience into these two unique segments, you can start bringing open rates back as a useful metric if you’re still attached to the process.
6. Go back in time
When in doubt, the Marty McFly approach won’t fail you. While we’re still in the spirit of unorthodox solutions, try this one on for size. Visit your past data, and create an audience segment filled with engaged members of your mailing list – important note, make sure that this list does not extend past August 2021.
You can then use this segment as a test group to judge the efficacy of your new subject lines/campaigns/templates/strategies etc., knowing that you’ll more than likely get an accurate result in terms of metrics. Make sure to take a deep dive into your CRM and look at the cross-channel user history to discover your key players.
Listen to your core audience!
7. Redesign your welcome process
As discussed previously, the change in the measuring of open rates has had a number of knock-on effects, one of which has been the ability to correctly maintain the health of your mailing list.
How can we maintain deliverability scores and ensure that our user engagement metrics are accurate? By being stricter with our process.
The pruning of an email list is only half the answer, and we’ve been focusing on this because it has been an easy and stress-free way of dealing with the issue.
But instead of just tempering with the output, we should take a closer look at the input.
By focusing our attention on who we bring on board, and the exact process in which we do it, we can avoid the sensitive issues that come up when it comes to dealing with who we have to remove and why.
A number of considerations are in order:
- Make sure to review your acquisition sources. An unwilling lead from a dubious source will do more harm to your mailing list than good. High-value, engaged targets are the name of the game here, and we shouldn’t settle. The changes that Apple made matter less if your audience is already engaged and shady acquisition practices will hurt you more now than ever before.
- Consider a double opt-in. Yes, it’s a bit draconian. Yes, you will lose a small part of your least engaged audience members. But it will solve more headaches than it causes in the long run, and besides, it’s best practice. We should only care for leads that are genuine about engaging with the brand.
- Finally, consider expanding your welcome journeys into multiple steps, and carefully monitor user activity. We know that user engagement peaks during the start, so it’s a good indication of how the user will react to the brand in the long run.
If need be, move less-engaged users to a “high-risk” segment and deal with them in an isolated capacity.
8. Improvise, adapt, overcome
Originally coined by the US Marines, and made famous by Bear Grylls, it’s an often-used quote, but with good reason.
It might feel like a cop-out to include this in this list, but hear me out. Marketing professionals are some of the most imaginative problem-solvers in any industry, so there’s no excuse here. EMS providers have given us a host of tools to work with, be it in terms of template creation, automation, segmentation, reporting, etc.
It’s time for out-of-box thinkers to shine – so grab a number of people from your organization that is known for their brainstorming potential, and have them find solutions that work for you. If the answers aren’t there yet, they soon will be.
This is the process. Working in the digital marketing industry involves often sweeping changes that force us to go back to the drawing board and reconsider everything we’ve done so far. It’s annoying, but it also keeps us on our toes and makes us better professionals for it.
The trouble here potentially is one of underestimation. It’s easy enough to consider this a unique situation that we can make workarounds for, and then promptly forget.
But a degree of future-proofing is in order here. If this is a sign of things to come – at least, if the recent rumors are any indication – and gets adopted by the likes of Google, then we want to make sure we are ready.
If you do the work now, you won’t get caught unprepared.