SKAdNetwork (SKAN) might feel like a novelty now, but soon will be bread and butter for marketers – a basic necessity for day-to-day operations. It’s easy to lose track of how we got here, so here’s a short refresher to prepare you for any SKAN updates that come after iOS15.
In the beginning
In late March, 2018, a truly terrible alternative install tracking solution was born, and it was christened ‘SKAdNetwork’. This was the same month that the Cambridge Analytica scandal hit the headlines, and only a few days after the FTC publicly confirmed suspicions that it was investigating Facebook over it.
Was SKAN born out of careful strategic planning? Or was it a knee-jerk response to increasing governmental pressures around data privacy? We will never know.
What we do know is that Apple had already started down the same road with Safari browser. Intelligent Tracking Prevention (ITP) 1.0 had been released in September of 2017, only six months before SKAN 1.0. It’s conceivable that the Cambridge Analytica scandal inspired Apple to effectively make a substantive public statement about data privacy in apps by releasing SKAN 1.0 ahead of schedule. Cambridge Analytica harvested the private data of 87 million users, and leveraged that data for targeted psychological interventions on social media intended to achieve political outcomes. They were deep in ITP territory, but maybe captain Tim felt like something had to be done about apps in real time. Orwellian is not a good look for a company that has the capability to see and hear everything you do, and track every movement you make.
Whatever the case, v1.0 was not fit for purpose. It was operationally useless and a lot of industry players didn’t even register its existence. Apple deepened its stance on consumer privacy, but did not release an update to SKAN until September 2020, over 2 years later.
SKAN 2.0 was announced at WWDC, in June of 2020. Of course, this announcement came as part of a set, along with the announcement that would shake the mobile advertising world: App Tracking Transparency. SKAN 2.0 was released, as expected, in September 2020, almost exactly one year ago, to accompany iOS 14. Version 2.0 included critical data parameters that were missing from version 1.0, including introduction of version number, source app ID, redownload value and conversion value.
Apple gave a reprieve, though, on the end of the world. They announced that ATT wouldn’t kick in until 2021. It’s a good job they did that, too, as 2.0 was… problematic. The privacy thresholds and time delays in particular made meaningful attribution almost impossible. When advertisers realised that Apple was not enforcing its fingerprinting policy, the vast majority fell back on MMP fingerprinting for attribution.
SKAdNetwork v2.1 was released in December of 2020, with a new public key for install validation. Then, in January of 2021, on Data Privacy Day, Apple published ‘A Day in the Life of Your Data,’ and confirmed that ATT would be coming in ‘early spring’ of 2021. That same day, Tim Cook addressed the Computers, Privacy and Data Protection conference in Brussels. He took direct aim at Facebook’s approach to data privacy, warning of an imminent ‘social catastrophe’ from the data industrial complex. That day, Apple also clarified its position on third party fingerprinting and announced coming improvements in SKAdNetwork, with support for view-through attribution.
Up to this point, the approach had been strong communication with limited action, but since version 2.1 it’s been clear that Apple means business.
Springtime for SKAN
It wasn’t until spring of 2021 that SKAN updates started to get interesting. In April of 2021, the protracted release of iOS 14.5 began. The dreaded ATT became a reality (albeit very slowly). SKAN 2.2 was released and the industry got some insight into the attribution logic. A new parameter was introduced: fidelity, allowing advertisers to distinguish between view-through and click-through impressions. We also learned that SKAN counts only the last 15 impressions. Bummer.
Shortly after, in May, iOS 14.6 was released (and SKAN 3.0 along with it). With iOS 14.6, SKAN postbacks were now relayed through Apple servers, obscuring the user’s IP address. We did find out, though, that the header will still include language. A new parameter, did-win, was also introduced in SKAN 3.0, allowing up to five other networks to receive a postback in addition to the one that won the attribution.
At almost exactly the same time, Apple made an undocumented (but significant) change to privacy thresholds, lowering them particularly for conversion value ID (less of an impact was seen in source app IDs). This behind-the-scenes change had an immediate impact on all versions, resulting in a substantial improvement in visibility of conversion values for many advertisers.
What lies ahead: iOS 15 and beyond
In June of 2021, Apple announced iOS15 privacy features, including the ability for advertisers to receive SKAN postbacks too (meaning SRNs won’t be able to mark their own homework any more). iCloud Private Relay (and IP address blocking on web) and App Privacy Report were also announced, indicating ongoing enhancements to consumer privacy features for iPhone users.
In August, Facebook announced that it will not send SKAN postbacks and is deprecating Advanced Mobile Measurement in October. This will have serious implications for advertisers, who will no longer have access to granular data for attribution, though for now Facebook will still allow MMPs to continue receiving it.
iOS 15 was released on Monday (September 20th). Unexpectedly, Private Relay is OFF by default in the released version. This is good news for web advertisers, as Private Relay previously defaulted to ON in the last beta version.
What lies ahead? If ITP serves as any example, this is the beginning of the app side of the global transition to a privacy first internet. We are closer than ever to the ‘cookieless future.’ Whether ITP was the catalyst or just a side effect of this transformation, Apple was first. While SKAN might not play out in exactly the same way as ITP, it’s safe to acknowledge the powerful forces pushing the market toward enhanced user privacy, and away from the commodification of third party data.
We hope and believe that Apple will continue to improve SKAN to make it more usable for app advertisers. Though we continue to see changes to SKAN, it’s disappointing that Apple seem to be over-investing in nice-to-have refinements (e.g. attribution assists) while still failing to adequately support the really important, basic needs of advertisers (e.g. geo country information or creative ID).
We also believe that Apple is likely to enforce its policy on third party fingerprinting. Failure to do so after aggressive positioning around app tracking and privacy could have undesirable consequences on the brand and result in a hit to hardware sales. They have, over the last several years, made trust the foundation of the relationship they have with consumers. Violating that trust might be a very bad idea, and could not only make winning over Android users more difficult but also provide ammunition to eager competitors.
On the other side of the equation, though, the App Store generates a significant proportion of Apple revenue at almost $74 billion (out of $274 billion in 2020). And, as we learned in the Epic court filings, the App Store operates at almost 80% margin compared to just over 32% on your iPhone. How much will a change in performance advertising impact App Store sales? It will be a careful balance to strike.
At Dataseat, we believe that companies that work with the changing privacy landscape will get ahead of the curve. That’s why we have made it our mission to provide gaming companies the tools they need to adapt to the post IDFA world.
Get in touch to learn more about our context-driven UA and in-house cross promotion solutions. [email protected]