SKAdNetwork 4.0 has already been announced, but with no implementation timeline as yet. This is the inevitable future of the industry as it moves towards greater confidentiality of user data. Many marketing experts and mobile app owners are getting ready for the market’s new challenges even as we speak.
In this article, we are going to:
Since iOS 14.5, all mobile apps have had to request user consent via App Tracking Transparency (ATT) in order to obtain IDFA, a unique device identifier that is required for personalized ad displays, targeting settings, and ad campaign evaluations and optimizations.
Today, this is something we have grown used to, but at the time it was a major step towards securing the confidentiality of user data.
To find out more about IDFA and how ATT works, check out our article MyTracker Solutions for iOS 14.5 Update
When users did not consent to sharing IDFAs, Apple offered its own framework, called SKAdNetwork. With it, Apple did the attribution itself and sent install information to an ad network without disclosing IDFAs.
SKAdNetwork was created for confidentiality of user data and attribution for ad campaigns without disclosing any device or iOS 14.5+ user information.
The first version of SKAdNetwork was released by Apple in March 2018, along with the iOS 11.3 release. Later, Apple made available version 2, which added the option to indicate conversion value, followed by versions 2.1 and 2.2 with the support of view-through ads. Finally, version 3 of the framework was released, with postbacks sent to multiple ad networks at once, indicating the one that delivered conversion. The list of versions and changes is available in Apple’s documentation on SKAdNetwork.
With SKAdNetwork 3.0, the SKAN mechanism could be described as follows:
For detailed information on how SKAdNetwork works, go to our documentation.
In addition to the complex random timer mechanism, marketing experts also faced other challenges when dealing with SKAN 3.0.
With SKAdNetwork 3.0, you can get only one postback once an app is installed. Moreover, user activity measurement is limited to a single time window within the first 24 hours. This means that if a user makes a conversion action (purchases a product) within 25 hours after a conversion is registered, this action will not fall into the 24-hour measurement window. Marketing professionals will receive a postback without a conversion value, which makes it impossible to reliably assess the effectiveness of an ad campaign.
Under Apple’s rules, a postback can contain the conversion value and an original app identifier only if sharing of these data meets the privacy threshold. While Apple does not disclose precise numbers, there is an opinion that it will require at least 88 installs per day per ad campaign. As a result, marketers have to pool ad campaigns or increase the campaign budget in order to reach the privacy threshold and get the treasured conversion value data.
In 2022, Apple announced and released the new SKAdNetwork 4.0, a version that enables users to get more helpful info while maintaining user privacy.
This time, SKAdNetwork was greatly improved and saw a number of changes that could be market disruptors, including:
For an overview of the announced changes, check out our article: Everything You Need to Know About SKAdNetwork 4.0
Today, SKAdNetwork 4.0 remains a technology that has been announced and is available but has yet to be implemented by market players.
For the industry to start switching to the new version of postbacks more quickly, there must be support from the existing market players.
According to Apple, in order to be eligible to receive a version 4 postback, all participants must meet the following conditions:
This means that it will take some time for the industry to feel the impact of SKAdNetwork 4.0. However, you can already start thinking about ways to put the new opportunities to good use.
The industry's move towards treating user data as confidential, the emergence of the first tools like SKAdNetwork, ATT, and the recently announced Privacy Sandbox Beta on Android all indicate a trend that is here to stay. There is good reason to start looking into and implementing tools like these in your projects.
We recommend finding out how SKAdNetwork works and drafting a strategy, even if you have no plans to use the framework right now.
To protect user privacy, SKAdNetwork 4.0 has four crowd anonymity tiers that determine the level of data granularity each postback will get.
The following Apple image illustrates how these tiers work:
Apple takes into account the crowd size associated with the app or domain displaying the ad, the advertised app, and the hierarchical source identifier the ad network provides. The system computes the volume of data to be included in the postback.
The anonymity tier determines the data you receive in postbacks as follows:
What does this mean for you? With SKAN 3.0, marketers had to consolidate campaigns to reach Apple's privacy threshold, while also getting conversion values in postbacks to evaluate the quality of traffic. While privacy thresholds remain in place in SKAN 4.0, it does not define exact values to be met. This means campaign consolidation will still be relevant.
At the same time, new privacy tiers make it possible to get coarse conversion values with the possibility of low, medium and high rather than a useless null, even if your campaign belongs to, let's say, Tier 1 instead of the top Tier 3 (fine-grained conversion mapped at 0–63). Encoding conversion events to the low, medium, high value will help you gain some insights into the quality of traffic.
In addition to a fine-grained 0–63 conversion value, SKAdNetwork 4.0 has introduced a coarse grain value (low, medium, high), providing a bit more detail in postbacks if the required privacy threshold is not met and a fine value is unavailable.
What does this mean for you? Don't neglect coarse conversion values. We recommend mapping events and revenue to these values right now to be able to assess user engagement at a later stage of app usage and receive data even if your campaign did not meet the privacy threshold.
Before SKAdNetwork 4.0, postback sending was pretty complicated, with only one postback returned for a set conversion value and a 24-hour waiting period needed to update the conversion value.
The new version offers three measurement windows and three postbacks, along with Apple's lockWindow method that will help you receive postbacks faster.
The first postback measurement window starts after the app opens for the first time (0–2 days). In 2 days, SKAdNetwork launches a random timer (24–48 hours) and sends the first postback with a set of data corresponding to the achieved crowd anonymity tier.
Postback 1 contains the following:
|Crowd anonymity tier|
|Tier 0||Tier 1||Tier 2||Tier 3|
|Conversion Value||-||Coarse conversion value (low, medium, high);||Fine conversion value (0–63);||Fine conversion value (0–63);|
|source-identifier||2 digits||2 digits||2–4 digits||2, 3 or 4 digits|
The second and third postbacks allow you to measure in-app activity from day 3 through to day 7 and from day 8 through to day 35 after the first launch. When measurement windows expire, SKAdNetwork will send postbacks with a 24–144 hour delay.
Postbacks 2 and 3 may contain the following:
|Crowd anonymity tier|
|Tier 0||Tier 1||Tier 2||Tier 3|
|Conversion Value||The postback is not sent||Coarse conversion value (low, medium, high);||Coarse conversion value (low, medium, high);||Coarse conversion value (low, medium, high);|
|source-app-id or source-domain||-||-||+|
|source-identifier||2 digits||2 digits||2 digits|
What does this mean for you? You can measure user activity over a longer period of time. SKAdNetwork 3.0 postbacks would return a nulled conversion value if the user was inactive within the 24-hour measurement window. In the new version, the first postback will now measure how active the users were within up to 48 hours, while the second and third ones will provide more insights into user activity in days 3–7 and days 8–35 after the conversion is registered.
For marketers to receive postbacks faster, Apple has introduced a new lockWindow method. This function makes it possible to finalize the conversion value and lock the window of the postback without waiting for the measurement window to expire.
What does this mean for you? Apply lockWindow to draw insights from your conversion values faster. To optimize ad campaigns, you need to grade acquired users as soon as possible. With lockWindow, you can get conversion values and validate user quality faster without waiting for the attribution window to close.
If you want to get ready for transitioning to SKAdNetwork 4.0 or opt for MMP to measure campaign performance for iOS 14.5+, MyTracker will equip you with a set of tools needed for SKAdNetwork 3.0.
Like other market players, MyTracker is now upgrading its solutions to support the new SKAdNetwork. To stay tuned, sign up to our Telegram channel.
MyTracker offers solutions to work with SKAdNetwork data, meet Apple requirements and continue collecting statistics for iOS 14+ devices.
With the SKAdNetwork dashboard, you can assess SKAN conversion dynamics by ad partner, and monitor conversion by source, geography and conversion value. This tool will enable you to grade users from each partner and rebalance the budget.
In case of insufficient data, you can always go to the Report Builder to add new dimensions and metrics. For instance, you can check SKAN conversions by SKAN version to find out which version of SKAdNetwork postbacks you receive and monitor your migration to SKAdNetwork 4.0.
As user privacy is taking center stage in app marketing, market players are attempting to balance privacy and accuracy of data. SKAdNetwork is a good example of these efforts, as with each new iteration it offers more tools to analyze anonymized data.
SKAdNetwork 4.0 will bring about a lot of important market changes which you must be prepared for as soon as possible:
At MyTracker, we are getting ready for SKAdNetwork 4.0 by developing solutions to work with anonymized data right away. Read more about our privacy-driven iOS 14.5+ solutions here.