Users have long protested against the collection of their data by all types of companies in order to generate money as well as improve their user experience within apps.
In this context, Apple has started the path towards greater protection of user data. This path has materialized with its newly released iOS 14.5 operating system, which, among others, brings news relevant to consumer privacy. With this new version, users will have to give their express consent (opt-in) to share their personal identifiers with brands.
One of the most relevant so far has been the so-called identifier for advertisers or IDFA, present in iPhones from iOS 6 in 2012 and is a unique key assigned by Apple to each device.
The IDFA was used by advertisers to measure and identify how users interact with advertising campaigns, app installs and in-app activities. If users did not want to share their IDFA, they had to explicitly choose not to accept it.
When the user first started his iPhone, he accepted the terms and conditions of use and, among the clauses, there was the possibility for advertisers to access that IDFA number, thus giving them the possibility to "track" their behavior to offer targeted advertising campaigns.
With the arrival of iOS 14.5, Apple introduced the App Tracking Transparency (ATT), changing the rules of the game for advertisers. ATT requires that all users choose to proactively share their IDFA and that developers display a notice asking the user to accept or decline.
While this is a positive move from a privacy standpoint, it will pose a challenge for many companies turning to mobile technology.
This creates a reality in which two completely different measurement methods coexist. Users who accept advertiser “tracking” will be grouped with the much larger segment of Android users, as the Google Advertiser ID (GAID) match remains unchanged. For those who choose to keep their privacy intact, the SKAdNetwork will be used, in addition to other tools.
Apple's SKAdNetwork launched in 2018 but was underused. However, now, with this new version of the operating system, SKAdNetwork will serve as the primary attribution mechanism whose goal will be to measure application installs and campaign performance without relying on user-level data.
Although SKAdNetwork offers several benefits, such as preserving user privacy and using a deterministic model, it also has several limitations. AppsFlyer conducted research before the iOS 14.5 update went into effect which showed that SKAdNetwork only captures around 68% of installations driven by non-organic activities.
"SKAdNetwork does not measure in real time, but measures with a 24 hour delay," said Daniel Junowicz, AppsFlyer Regional Vice President for EMEA. “In addition, the way to measure a user's quality has been limited by the new Apple framework which limits the way to cultivate and enrich quality information. In addition, the fact of not being able to identify the user limits retargeting ". Therefore, marketers should not consider SKAdNetwork as the only form of measurement.
Marketers are looking into new acquisition channels that complement SKAdNetwork to get a complete picture when their access to user data is limited. In an effort to maximize user privacy while delivering a good experience, a combined approach is required that leverages user-level data (provided the user gives their consent), SKAdNetwork requests, learning automatic and aggregate attribution. In this sense, advanced solutions are being developed that focus on aggregate measurement, machine learning and incrementality tests. Marketers are also placing more emphasis on web-to-app as a marketing strategy.
- Machine learning and predictive analytics
One of the biggest challenges advertisers will face with SKAdNetwork is the timing mechanism, whereby the measurement will be limited to the specific activity that occurs in the first 24-72 hours. How will advertisers make critical decisions about their campaigns based on such limited data? The answer lies in predictive technologies that will allow advertisers to leverage early engagement signals in the first 24-72 hours and thus predict campaign performance over the long term. This revolutionary advance will become a powerful tool in any advertiser's arsenal.
While traditional models “match” an ad click or impression with a conversion, incremental solutions use test and control groups to isolate many affected variables and help marketers optimize sales growth. As a result, marketers will be able to understand the real impact of their investments by finding out which conversions are the result of marketing efforts and which would have occurred organically.
- Streams from the web to the app
Web-app flows take a user from a web page to the corresponding application. The mobile web is increasingly seen as a crucial point of contact, as consumers use it to learn about a brand, its offer and what it can offer them. Advertisers acknowledge and respond accordingly. After all, web advertising is cheaper and with tools such as banners and deep linking (deep link), advertisers can "stretch" the consumer's journey to the last point of contact, their application, and thus create a better experience. user internally and ultimately a higher conversion rate.
The goal is unanimous: to privilege the user's privacy before anything else and to mitigate previous practices relating to the processing of his data. However, with Apple's SKAdNetwork significantly limiting the data available to an app developer on iOS, the challenges are multiplying.
"There is a lot of uncertainty on the part of advertisers in terms of ways to promote their apps in the new iOS landscape, which could mean that the budgets allocated to this system are decreasing," says Junowicz, who encourages advertisers to "try how these new options as soon as possible to know what to expect from this paradigm shift ".
This makes a robust mobile attribution platform especially essential to address this new age of privacy. This platform will allow marketers to continue measuring the success of their campaigns with user-level data as long as they provide their consent or by combining other measurement solutions.