Preparing for Apple’s Removal of Email Tracking Pixels
A couple of decades before the rise of easy web browsing, Ram Tomlinson invented email. It has remained one of the most efficient and profitable means of communicating with an audience. It is also the foundation of some of the most successful media companies.
Email works. Unlike most other mediums, email gives publishers the ability to go directly to the reader. It’s why an email subscriber is far more likely to buy a paid subscription than someone visiting from social or search. Email is a distraction-free communication that feels one-to-one.
It’s also become a very lucrative monetization opportunity for ad-driven media companies. When running native, well-written, direct ads, the revenue can be very good. Users react better to this—multiple times better—than your standard IAB banner ad.
That’s why many in media were concerned when they saw the news from Apple that in its upcoming operating systems, it would be introducing a feature that blocked open rate tracking. According to the press release:
In the Mail app, Mail Privacy Protection stops senders from using invisible pixels to collect information about the user. The new feature helps users prevent senders from knowing when they open an email, and masks their IP address so it can’t be linked to other online activity or used to determine their location.
I’m all for privacy. I think tracking has become a bit too intense and, honestly, I can’t fault users for not wanting companies to know what they’re doing. As a society, businesses have abused their customers’ privacy.
However, the publisher and the user are both going to suffer because of this change. Let’s dig into that a bit because it can help prepare how we should respond.
The big change for the user is that they will have to opt-in to share that data with the publisher. As we have seen with app tracking on iOS, under 10% of people do that. Therefore, we shouldn’t expect a bunch of willing participants.
There is a reason why this a negative for users, though. When your newsletter starts to get really big, one of the most important things you can do is clean the list. Rather than waiting for people to unsubscribe, we, instead, opt to actively remove people that are just not opening our newsletter. This is necessary for big senders. I wrote last year:
But there are lesser-known problems that many operators might not know about that occur when your engagement drops.
Two, in particular, jump out and they’re related.
The first is a poor sender reputation. Internet service providers assign a score to each sender to determine how reputable we are. That score takes time to build and it can slowly deteriorate as time goes on.
The second is the domain blacklist, which keeps track of bad actor domain names. The web administrator can use these blocking lists to help prevent spam from getting through to their user emails.
Essentially, deliverability depends on your newsletter having high engagement (being opened is one part of that). If publishers don’t have open rate data, how can we accurately remove the right people? With this privacy change, a user using Apple Mail could be opening A Media Operator every single week and I wouldn’t know, so I’d have to churn them.
That’s obviously not what we want to do and just delivering a blanket message to subscribers that they need to let us track them so they can keep receiving our newsletter is not really an option.
So, what can we do?
Most importantly, we need to look at click data rather than open data as the primary engagement metric. By doing this, we can theoretically extrapolate opens. If a user that doesn’t show as having opened a newsletter has clicked something in the newsletter, we can assume they opened. My guess is that ESPs will build the tools over the coming months that track this for us.
But the question remains: how do we make sure people click something so we know they want to remain subscribed?
When a user signs up, it’s normal for them to receive a welcome email series. This might include things for them to click or it might just be information that they need to know about what to expect and additional products you offer.
However, once that welcome series is done, we don’t do much to make sure the user is still enjoying our product. That needs to change.
Every few months, publishers should think about sending a “do you still want to receive this newsletter?” message to all subscribers. Throw a big button in it that asks them to click if they still want to receive it. That click will tell you that the user is engaged even if you can’t see if they’re opening it. This double opt-in is their way of telling you they still want your communication.
This is not too dissimilar to the types of emails we send when trying to renurture prior to churn. The difference is that more people should respond since this also includes active participants rather than just the people who are not opening.
Increase clickable opportunities
A simple way to start getting clicks in the newsletter is to actually include things for people to click. AMO is the type of product where it is meant to be read right in the inbox. That’s probably not the smartest thing for me to do when the open rate, alone, is not efficient enough.
But there is a simple way to get a click. Ask for it.
At the end of every newsletter, ask the user what they thought with a simple survey. Have a thumbs up and thumbs down that click to a different URL. You’ll learn two things here actually. First, who opened the newsletter, which we want. Second, if they liked it. Sometimes a little reader feedback—even if it’s not a reply—can help inform how we’re doing.
When all else fails, make a game out of it. Back in February, Digiday did a piece on how Bustle was planning to build a huge newsletter business. And this part jumped out to me:
Bonner added that one strategy for encouraging engagement, and thus higher CPMs, is Inverse’s proprietary rewards program that is currently only operating within Inverse’s newsletters. It acts as a “motivational incentive” for subscribers to track how many days in a row they open the newsletter, Bonner said. If they open it every day or a month, for example, they are then entered to win a prize, often from one of the newsletter’s sponsors.
Oshinsky said that while a rewards system can definitely help increase open rates, being able to translate that engagement to proof of a base of “super users” may not be as straightforward when dealing with informed advertisers.
I poo-pooed on this. As I wrote in my newsletter:
You can’t gamify engagement. It either exists or it doesn’t. If the incentive is to get users to open the newsletter to win a prize, BDG will absolutely drive a higher open rate. Humans like winning things. But the big metric that many advertisers look for from newsletters is clicks. It’s a direct response business. Therefore, if the readers are not engaging with the ad, it doesn’t matter how many opens you get.
I stand by that statement, but we are now looking at trying to ascertain opens rather than artificially inflating the number of opens. It’s about messaging to the audience. While they’re related, they’re not identical.
In this case, I think what Inverse is trying to do could be replicated. When a user opens a newsletter, if they participate in the previously described survey, they receive a point. At the end of the month, those points are randomized and a winner is picked to receive some sort of a prize.
It’s a little bit of everything
Ultimately, between reminders, polling, gamification, and just having links that are worth clicking, we should start to understand better what percentage of people are actively opening the newsletter even if we can’t track it.
The click is now going to be the most important metric to help us backdoor into an open rate. If your ESP is not already working on something to help give you an estimated open rate, you should inquire about it.
But what about media companies that sell ads in newsletters? Without an actual open rate, aren’t they in trouble?
The types of advertisers that tend to spend in newsletters are those that are more actively tracking engagement with the unit. Brand advertising matters, but clicks matter even more. Therefore, nothing has really changed here.
Pricing might get a bit more complicated, though. However, the way some publishers might determine rates is moving from a cost per open to a cost per click and then backdooring into a price. If an ad gets 1,000 clicks, on average, and your CPC is $10, then the unit should be sold for $20,000 versus if it gets 1,000,000 opens and your CPM is $20. The price is the same, but you’re pricing on more actionable data.
I should caution, though… In an effort to get the click, many publishers will go too far and try to trick users into clicking. And let me just say that the ultimate reason an advertiser wants to spend money is that the product is actually working. Artificial clicks that create an inflated open rate won’t change the fact that the product is bad. And even if you can talk about a high open rate, without the user wanting to engage, nothing will matter.
We need to intelligently use click data to come up with strong estimates about how engaged our audience is to help with sales, but that it can’t replace a good product. We need to understand that quality trumps everything. If we do that, our businesses will be in a fine place.