Apple is about to make a big change to how digital advertising on mobile devices is managed. The new version of its operating system, iOS14, will increase user privacy at the expense of how marketers target audiences and measure their campaigns’ performance.
To measure the impact of in-app advertising for driving app conversions (an ‘app-to-app’ customer journey), Apple has announced a new attribution solution called SKAdNetwork. It will provide advertisers a slimmed down way of measuring campaign performance even where people have asked for their data to be kept private. Essentially, the data will be delayed, aggregated, and limited. You can read more here.
What about marketers with an app-to-web customer journey?
Measurement of website performance does not rely on the same technology as measurement in apps. Whereas app measurement relies on an ‘Identifier for Advertisers’, the web (including mobile web) is measured using a range of tools including cookies and UTMs.
So if you are advertising ‘in-app’, such as with Instagram, Twitter or YouTube, and sell your products on your website, are you in the clear?
Well, no. Apple has realised that by using the existing methods, marketers would be able to easily link their in-app advertising with new website users and customers – even if the person had opted out of tracking on the referring app. Very much against Apple’s new ‘rules’.
Private Click Measurement
So, to ensure the privacy of its users, Apple has introduced a new protocol for measuring how people interact on websites: ‘Private Click Measurement’ (PCM).
PCM is designed to restrict the amount of data from a web user is shared across platforms, but in a way that can offer marketers some insight into their performance.
It purposefully prevents advertisers from feeding back user-level information to the app that delivered the traffic. This has the drawback of no longer helping the platforms’ algorithms to optimise towards valuable users (as they will no longer know what has happened after the click).
PCM will only allow the referring app to know that some people clicked an advert, and at some point in the future some people converted. Note this is very different from determining that this person clicked on an ad and then this same person converted. Apps won’t be able to link the two events together to one specific user.
With PCM, all users will be treated like they are in private browsing mode. And any data made available for reporting will be aggregated and delayed (similar to opted-out app users with SKAdNetwork).
(As a side note, advertisers will still be able to use UTMs to get more insight into their campaigns. This information just won’t be able to be passed back to the platforms.)
Clearly, this is a major setback to digital marketers looking to continue understanding and improving the value of their mobile advertising.
Facebook and ‘Aggregated Event Measurement’
Because PCM is designed to prevent apps such as Facebook from determining the value of the different campaigns run on its platform, you can imagine that these platforms aren’t too happy with the new restrictions.
In response, Facebook is releasing ‘Aggregated Event Measurement’. Or AEM for short.
AEM will help Facebook and its advertisers to comply with the requirements of PCM, while improving its ability to attribute (and learn about!) app-to-web conversions. It supports the measurement of web events in iOS14 users. Data available for measurement and attribution will still be delayed, aggregated and limited.
There are going to be a range of changes to how marketers run app-to-web campaigns on Facebook properties.
First, these campaigns will be limited to tracking eight post-install conversion events. This is more restrictive than the app-to-app changes (where there is a limit of 64 events). These eight events will need to be ranked in priority order, and only the highest-ranked event that the user completes will be attributed to the campaign.
For example, imagine an online retailer. They might define their eight events as below. The furthest a person may go is to the fifth. Only that fifth ranked event (‘Add to cart’) would be reported back on. We would have no way of knowing for certain if the four lower ranked events also happened.
If you are doing Value Optimisation campaigns in Facebook, you will only be able to track four conversion events, not eight. And because the campaign data is aggregated, optimising towards revenue metrics such as ROAS will be less accurate too.
With the data being aggregated and limited, we will also lose the ability to dive into the demographic breakdown of conversions in Facebook. So you’ll no longer be able to say 60% of your customers from Facebook are ‘women aged 29-34’, for example.
Finally, Aggregated Event Measurement only supports a post-click attribution window of 1 day. In the past we have been able to track, measure and optimise activity based on 28 days of post-click or post-view conversions. These attribution windows are being deprecated. Expect to see Facebook ad performance weaken as you lose the ability to ‘claim’ any slower conversions for your campaigns.
What do these changes mean for digital marketers selling goods on their website?
We think there could be a move back to measuring the longer term impact of campaigns, rather than the immediate next-day ROI it delivers. Short term metrics are going to show digital advertising underperforming and the budgets hard to justify. Digital marketers need to get comfortable thinking about establishing sustainable growth, rather than short-term gains.
162-168 Regent Street
+44 203 096 5108