Wrong Assumptions of SKAdNetwork and The Reality After the Loss of IDFA

When Apple decided to take the bold step of standing up for their users’ data privacy by taking back IDFA (ID For Advertisers), mobile app marketers thought all was lost.

The revoking of IFDA started with the introduction of iOS 14 in September 2020. The implication is that iOS 14 users will receive a one-time prompt via a text box controlled by Apple that explains all the data access that the App Publisher is requesting the user to allow.

SKAdNetwork was introduced to replace the attribution after the loss of IDFA, and while this proved to be a challenge to mobile app marketing, some app marketers do not fully understand how SKAdNetwork works.

Several wrong assumptions have been made by app marketers who try to understand SKAdNetwork. We’ll explain these wrong assumptions in this article in a bid to help you understand SKAdNetwork and plan for a more robust mobile app marketing strategy for iOS devices.

Losing IDFA for Marketing Teams

Losing IDFA meant a lot to mobile app marketers, especially on the user-data level. The user-level data, which includes GEO, Device, App Version, In-App Event and Revenue Data, etc., is used by growth teams as one or more of the following:

• Attribution – Mobile Management Partners (MMP)
• Analytics – Performance analytics, creatives, ad monetization
• User Experience – Deeplinks / Web to app
• Fraud Prevention
• Campaign optimisation / Targeting
• User list – RTG / Lookalike / Xpromo

So, without IDFA, what it simply means is that mobile app marketers will be unable to fully track their new users within the app funnel, and have trouble calculating their revenue. They will have a situation whereby they decide how much to spend on different mobile ad channels but can’t calculate their revenue or return on ad spend (ROAS).

With SKAdNetwork, the chances of the users accepting to be tracked are very slim. Even though you can finetune the message on the popup, the message passed across to the user remains the same, “you want to track them.” Very few will accept that.

So, the challenge now is for marketing teams to rethink how they track, onboard and get the data they need to effectively advertise as user-level data will no longer be available.

Assumptions of Advertisers About SKAdNetwork

Because SKAdNetwork poses uncertainties regarding how to use it in day-to-day marketing decision, lots of wrong assumptions flying around concerning the innovation. Here are some of the wrong assumptions:

1. Timer

Timers, extensions, and delays are some of the misconceptions of the SKAdNetwork. SKAdNetwork has a 24-hour timer that starts when the SK function is called for the first time – which is on the app first launch. Once this timer expires, the conversion value is locked in.

You can play around with this, however. You can reset the timer whenever the user completes an event within 24 hours. Consequently, if the user does not engage with the app for a day, or does not generate a new event down in the funnel, the timer will not reset, thus ending the loop.

2. Change

Server-based solutions for SKAdNetwork allow advertisers to change the conversion value configuration in the cloud, where the command will be sent instantaneously to the SDK. This is to keep the customer logic updated with the desired configurations.

With SKAdNetwork, however, each configuration change triggers a short period that makes it impossible to determine whether incoming post-backs relate to the old configuration or the new one. This creates an extra “noise” in the measurement.

To minimise this noise, you should not change this configuration frequently, not more than once every two weeks. Alternatively, to walk around this, you can dedicate 1 of the 6 bits of the conversion value to indicating whether the post-back relates to the old configuration or the new one.

3. Server Events

A cloud-based solution can forward events to the client-side, changing the conversion value accordingly. The limitation to this, however, is that if the server-based event was triggered before the SKAdNetwork timer expires, the end-user must reopen the app again before the timer expires for the event to affect the conversion value.

Because this type of flow works for client-based events, you have to stick to client-generated events where possible.

Gaining Opportunity out of SKAdNetwork

The implication of SKAdNetwork to mobile app marketing is that marketer would not be able to measure “everything” or optimize their campaigns based on “everything”. However, all hopes are not lost. You can still measure, optimize, and excel in your mobile app marketing strategy by proper planning, testing, and validating assumptions.

Optimisations will be on more on aggregated levels instead of user level.