Over the last few years Apple has released many instances of privacy preserving protocols across its platforms. With this latest update in iOS 14, it’s a good time to assess what’s been happening, and how it impacts our advertising campaigns. We’ve included some recommendations in the summary, and we’ll follow up this piece with more guidance on solutions, as the landscape evolves and we learn more about the impact of these changes.
Apple introduced its ATT (AppTrackingTransparency) framework for iOS 14, which includes a new prompt being seen by users on iOS devices when they download an app. Here’s what the prompt looks like:
This is already being seen by some users, and the industry expects this to be fully rolled out to all users by March 2021.
What changes with iOS 14?
If a user accepts, not much changes. The IDFA, which stands for Identifier for Advertisers and is essentially a unique ID for each iOS device, will continue to be shared as it previously was. But the industry expects the majority of people to decline this prompt, meaning the IDFA will no longer be available for advertising purposes.
It’s useful to think about this in two ways, firstly, how this impacts App Advertisers (advertisers driving users to their app), and secondly, how this impacts Web Advertisers (advertisers driving users to their website).
For App Advertisers, Apple will use what it calls SKAdNetwork to enable the attribution of events or conversions back to an advertising click. There are a few protocols in place, such as a limit to conversion events (64), the number of campaigns that can be used for an iOS app (100), and a random timer to further obfuscate the connection between the event and an individual user. These protocols allow advertisers to see that a campaign is driving conversions, but not that a specific user made that conversion. We can expect the MMPs (mobile measurement partners), which most app advertisers will be using, to be building this into their product.
Here’s a diagram showing how this works:
For Web Advertisers, we have two scenarios:
Scenario one: Someone clicking on an ad in the mobile app, such as Facebook, Pinterest or Twitter
For scenario one, when a user clicks on an ad in, let’s say Facebook, and they’re redirected to an advertiser’s website, Facebook will register the click. A new requirement in the Facebook platform due to iOS 14 is to select up to 8 events per domain, and in priority order, so that if the user performs an on-site action which is listed as one of the 8 chosen events, this event will be registered by Facebook. Only one event will be tracked, the highest priority event. However, while the ad click in Facebook is registered in real-time, the conversion may experience a delay of up to 3 days. This is where advertisers might see a message in the Facebook Ad Platform stating that the results are either delayed or being modelled.
For social platforms like Facebook and Twitter, the vast majority of engagement happens via the app, so the above is likely to be the case more often than not. The result is fewer trackable events, delays in reporting, and other backend implications these platforms will experience, such as impacted algorithms due to reduced data visibility.
Scenario two: Someone clicking on an ad in a browser environment, such as Google Search or a Display Banner
For scenario two, the main difference worth calling out is what happens on Safari, as the browser has slightly different protocols in place. We’re focussing on Safari as it has 90%+ usage on iOS devices.
Safari has released versions of Intelligent Tracking Prevention (most commonly referred to as ITP) over the years, which impacts how user data is collected and shared. In short, ITP ensures that only click-events are tracked from advertising campaigns, so no ability to track or attribute impressions and that click measurement is done in a way that preserves an individual’s data. Over time we expect Apple’s Private Click Measurement protocol to be fully rolled out, which can be viewed as the web-equivalent of the SKAdNetwork mentioned above. We can expect this to provide similar measurement opportunities for advertising across the web, including on Google Search and through Display Banners. In the short term, however, the impact of this latest change will be seen in the form of limitations in data available for measurement and optimisation, with campaigns such as Google App campaigns being heavily impacted.
A final piece of the puzzle, related to both app and web advertisers, is around the use of email lists. While it has been expected that email lists, or other first-party data assets, would still be usable in their current form for iOS campaigns, Apple has put further requirements in place to ensure consent has been granted via the ATT framework by the user. Although the GDPR is already in place to ensure all advertisers have received consent before using email for advertising purposes, the below excerpt from Apple’s developer site does imply that using these email lists in Facebook or Twitter, as Facebook and Twitter would be classified as an Ad Network, in this case, may still require ATT consent.
We see three core areas being impacted with all of these changes:
- An advertiser’s ability to measure advertising campaigns, due to conversion event limitations, changes in available attribution windows, and delayed reporting
- A platform’s ability to create effective models for targeting and optimisation across all platforms, due to the reduced data being seen across iOS devices
- The way audience activation is managed moving forward, due to reduced audience pools, limitations on granular audience insights, and limitations around features such as dynamic ads
The full impact of this will only be seen over time; there are, however, a few things we can do now to ensure we’re in as strong a position as possible.
Firstly, ensure the recommended or required changes from platforms or agency partners, such as domain verification, have been implemented ASAP. At Croud we’ve already shared these with relevant clients, so please reach out for any support.
Secondly, where appropriate speak to your account teams about implementing server-to-server solutions such as Facebook’s Conversion API or Google’s Enhanced Conversions, to gain more visibility on advertising campaigns. These solutions work in conjunction with website tracking pixels, and operate from a server (instead of the user’s web browser), supporting conversion tracking by providing an alternative measurement for the same user action.
And lastly, keep us updated with what you’re seeing internally with regards to these wider industry changes. There is unlikely to be a one-size-fits-all solution for all advertisers, and so finding the most effective approach to activating, measuring, and communicating campaigns and their performance is best done in a collaborative way.