The following post is a re-post from Adjust that we thought may be valuable for understanding the terminology out there.
Ever since Apple announced iOS14 and the AppTrackingTransparency framework (ATT) there has been a lot of confusion in the ecosystem about what is allowed — and what isn’t.
What’s important to keep in mind is that Apple’s goal with ATT is very similar to the purpose of privacy regulations like GDPR. The rules exist to allow users to choose whether a first-party can share their unique, identifiable, and persistent data with a third party.
Sounds simple, right? So why is there so much debate over what is covered by the rules?
Part of the confusion stems from a lack of a common language across the entire industry. A lot of industry players are using different terms for similar concepts.
One complicating factor is that the industry had been using ‘fingerprinting’ as a catch-all, encompassing both actual fingerprinting and methods of probabilistic attribution. With the upcoming 14.5 changes, some companies (Adjust included) moved from fingerprinting to strictly probabilistic attribution. This has meant picking apart what people understand by ‘fingerprinting’ and explaining what is still allowed.
I wanted to define a few terms that are important to understand and differentiate:
- What is fingerprinting? A method to track users cross-site by utilizing device information to create a persistent and unique ID. Some techniques used to achieve this fingerprint include capturing font metrics, using WebGL (and canvas) properties, combined with using certain hardware properties. This data makes the fingerprint persistent and uniquely able to identify a user. The main use of fingerprinting and fingerprint IDs is to track users across different websites and apps that would otherwise not share any common ID. For example, fingerprinting is used to create device graphs, which are clearly against Apple’s guidelines.
- What is probabilistic? As an MMP we don't track or target users across sites or apps. All we care about is attributing an install to an engagement with some degree of certainty. As 80% of installs happen within the first hour after click, such attribution does not require any persistent ID. We can make our predictions with temporary data that becomes obsolete within a few hours. Therefore probabilistic attribution for us is simply based on device entropy and patterns. We look at parameters such as time of click, time of install, and basic device info. These limited parameters allow us to estimate the source of an install for a few hours after a click.
- As an advertiser, why should I use probabilistic if SKAdNetwork is more accurate?
Probabilistic attribution isn’t meant to replace SKAdNetwork and will never be as accurate. However, it provides real assets to any advertisers running campaigns. With probabilistic attribution, your media partners will be able to optimize your campaigns, improve their models, and give you the best ROI.
- So, does that mean that I can share data with my media partner?
Yes, for instance, sharing a keyword back to the media partner that allows them to attribute the probabilistic install to a campaign is acceptable. None of the data shared allows for cross-site/app tracking or targeting.
- What is conversion modeling? Conversion modeling is extrapolating the behavior of consenting users to model the aggregate behavior of all users. There are two forms of this that we’ve heard are acceptable:
- For attribution purposes. Analytics companies will take consenting users and their behavior after an install, and use that data to apply similar metrics to all users. This will allow you to know your cohorted metrics such as LTV and ROAS. As marketers, data must be accurate, so you’ll always have to question how accurate this truly is. The accuracy of the conversion models will depend on your opt-in rate.
- For targeted advertising. Similarly, the media company would use the subset of consenting users to serve those who didn’t consent a relevant ad offering based on similar contextual signals.
- What is SKAdNetwork? SKAdNetwork is Apple’s attribution framework. It gives media channels the ability to have a source of truth on attribution in this new privacy-centric world. Its strength is that it provides deterministic attribution with almost 100% accuracy. We’ve tested it and SKAdNetwork comes within 2% of the accuracy of deterministic attribution via IDFA. Note that if a campaign was attributed 1000 installs with IDFA previously, SKAdNetwork may attribute 900 installs and 100 redownloads. That is because SKAdNetwork will only credit an install on an iTunes account once. This is why it is important to look at the sum of installs and redownloads to keep things consistent with the pre-ATT era.
It is also important to keep in mind that SKAdNetwork does not yet cover all available inventory, and publishers are still adopting the framework. Until coverage is close to 100%, this will also cause SKAdNetwork to report fewer installs than the amount that deterministic attribution via IDFA reported prior to iOS14.5.
- What opt-in rate can I expect? Since 25% of users globally block sharing of IDFA, 75% will be eligible to see the consent modal. From our analyses, we’ve seen ~40% of those users will consent to share IDFA, which will yield a ~30% IDFA rate on your app. Make sure to read our guest post on AdExchanger for more information on crafting a pre-permission prompt.
So who stands to win and who stands to lose? In every platform change, some people benefit more than others. Mobile companies who don’t understand the ramifications of these changes are set to lose out on a growth opportunity.
However, companies that move fast are in a good position to innovate and take advantage of the opportunity. Agile firms, that also have first-party data, are best positioned to win.
For more information on how Adjust will support your growth, make sure to check our latest guide, How Adjust is supporting growth through iOS 14.