The future of digital advertising will not be based on an ID
April 13, 2021
What will advertising, and particularly targeting on mobile, look like when Apple’s mandatory IDFA tracking changes take effect? (and if Google follows suit?) Pan Katsukis from Remerge gives us the answers.
A world without advertising IDs is difficult to imagine at the moment. And yet, many companies had already started to prepare for this scenario before Apple’s 2020 announcement that users will soon have to opt-in to app tracking on iOS14. The announcement of this change caused a stir among advertisers, developers, and advertising platforms such as Facebook. The latter publicly denounced Apple for feared losses in advertising revenues and a drastic reduction in tracking options.
After Apple postponed the tracking opt-in to spring 2021 - it should be mandatory from version iOS 14.5 - an update was written into the App Store guidelines. By now, those who will be affected by the upcoming change should be looking into ways of informing potential app users about the opt-in process and exploring alternatives to tracking and targeting. After all, Google could soon come up with a similar feature for Android. Google is one of the companies that responded quickly to Apple's news and adapted its own advertising SDK with an update at the beginning of September 2020. They also provided app developers with tips on how to deal with the mandatory opt-in changes.
"At the end of 2019 we already aligned our strategy towards a no-ID future"
Companies in the app marketing industry have been grappling with what an advertising future could look like without a personal identifier long before Apple’s big announcement. Among these international companies is Remerge, a programmatic app marketing platform that specializes in solutions for app retargeting, user re-engagement, and incrementality measurement and uplift optimization. That's exactly why we spoke to Remerge CEO and co-founder Pan Katsukis. We wanted to gain a deeper insight into the status quo of the mobile marketing scene and the expectations of professionals regarding the future of the industry.
The interview
Mandatory opt-in for mobile app tracking
OnlineMarketing.de: Apple's announcement regarding the mandatory user opt-in for tracking apps in the App Store has created a shockwave in the marketing landscape. Were you surprised by the development at Remerge?
Pan Katsukis: Anyone who has followed Apple's development over the past few years knows that the topic of privacy has been the focus of its corporate strategy for some time. It was clear to us early on that this would also affect marketing. That is why we decided to align our strategy towards a no-ID future at the end of 2019, and our team committed to it. At the beginning of 2020, before Apple's announcement of the new privacy framework, we started working on new products that can work in a world without an ID. When Apple announced its concrete plans for a stronger privacy-driven app environment in the summer of last year, we were well prepared and were able to quickly and effectively set new priorities.
Estimating revenue losses
In this context, is it possible to estimate how much revenue losses an average SME could actually face?
There are many factors that will play a role in the future: How will advertisers behave? Are marketing providers able to quickly provide them with good alternative products that also work without an ID? In order to estimate the consequences in terms of sales, one can of course draw a comparison with similarly drastic developments in the past, for example when Apple rolled out ITP for Safari in the web area. Without cross-site tracking, advertising revenues fell by 40 percent. However, Safari plays a relatively small role on the web, very different to the iOS operating system for app developers. The need to reach iOS users through advertising will definitely remain high, so more will be invested in alternative solutions. Accordingly, there are estimates of sales losses that are only in the mid to high single-digit range (four to eight percent). In the first few weeks there will certainly be a shock and the losses will be higher, but over time this should recover once the uncertainty has subsided.
The personalization of marketing activities
Facebook has created a dedicated website to point out the dangers that arise for SMEs in light of the IDFA update. In your experience, are these companies currently afraid that their marketing efforts might be less personalized?
I doubt it. In relation to this website campaign, it looks more like Facebook is waging a proxy war against Apple. Targeting, for example by city, will still be possible. What will change is the measurability of campaign performance, and as a result Facebook will no longer be able to simply attribute itself to itself, as in the past. Advertisers will rely on Apple's new SKAdNetwork for attribution - where there are 40 percent fewer visible conversions. If the SMEs see poorer performance, then they will also be more cautious with the budgets or shift them to other channels.
The impact on mobile attribution
Which areas would actually be affected? Is this only about targeting and retargeting or also about attribution models for the apps?
In fact, I think attribution will be affected even more than targeting. Apple now forces advertisers to use SKAdNetwork for attribution. MMPs and attribution platforms such as Adjust and Appsflyer must also use SKAdNetwork and must stop using fingerprinting or probabilistic attribution models. In comparison to today's models, SKAdNetwork has a relatively rudimentary design; the data is aggregated and only made available after 24 to 72 hours. This will, of course, make the timely campaign optimization more difficult.
Developing a strategy to measure and optimize advertising
What do marketers have to prepare for now - alternative solutions, an app optimization that is as compliant as possible, or both at the same time?
If marketers want to continue reaching iOS users, then they need a good strategy for measuring and optimizing advertising. The top marketers have been testing the no-ID inventory for a few weeks to find the right strategy. As already mentioned, SKAdNetwork measurement on its own is not enough to optimize advertising. I am sure incrementality tests will be used to achieve real-time results and gain scientific insights on the impact of a campaign.
It makes sense to be active in this area from the very start. Right now, prices for no-ID inventory are likely to be very cheap because many marketers will be unsettled and hesitant. Large web platforms such as The Trade Desk have also announced that they will not buy no-ID inventory, which means that competition for advertising space will be less than it is today.
Alternatives to app tracking
Are there any direct alternatives to app tracking that could determine a similar user data set?
There has been a debate over whether fingerprinting or probabilistic methods will be allowed, but Apple made it clear in its updated FAQ that doing so would risk expulsion from the app store. Apple clearly wants to put a stop to the practice of tracking and targeting individual users without their consent. Anonymization alone is not enough. So the basis of alternative methods must be data aggregation, that is, to combine users into groups and populations. This is done, for example, in incrementality tests. Here, the users are divided into a test group and a control group and the performance of the two groups is compared. The test group sees the advertising campaign, the control group does not.
« I am sure incrementality tests will be used to achieve real-time results and gain scientific insights on the impact of a campaign. »
Pan Katsukis
Acquisition and retention in a privacy-centric environment
How can apps continue to thrive in terms of acquisition and retention as we move towards a more privacy-centric environment in iOS?
With SKAdNetwork, Apple offers the basis for measuring campaigns. However, this only provides time-delayed and aggregated results and does not reflect retention values. Additional measurement methods such as incrementality tests are a useful component in order to see real-time data and retention values. The fact that inventory will become cheaper in the short term will also have a positive effect. This will make it possible to spread campaigns and still achieve good results. The focus will be on advertising media and context. If you keep this in mind, it helps with optimization.
The focus on aggregated data
How will user acquisition and retention likely change? How does the lack of data affect user-level targeting, creative personalization and attribution models? Do you have an example scenario for this?
Much of the targeting we see now will also be possible in the future. For example, we continue to use Dynamic Product Ads for personalized advertising. The basis of this, however, is not which specific articles users have previously viewed, but rather aggregated data such as the city in which they are located or the time of day. For example, a delivery service can display a suitable advertisement with a coupon for lunch in the city where the users are located.
Sustaining app growth through programmatic advertising
How are the demands on DSPs changing and what factors will be important in sustaining app growth through programmatic advertising?
We see three main factors: transparency, optimization and scale.
In the future, DSPs must be able to generate more knowledge in order to provide the foundations for making decisions about optimization. Relying solely on SKAdNetwork data will not be enough. Incrementality measurement would be an example of something that supports decision-making around optimization. Contextual data and strong creatives will also play a bigger role in optimization. By scale, I mean how many queries per second (QPS) the DSP processes and how intelligently it can provide insights that help deliver a relevant, high-performance range. More contextual data points on the supply side can therefore help with optimization.
« In the future, DSPs must be able to generate more knowledge in order to provide the foundations for making decisions about optimization. »
Pan Katsukis
Incrementality and true iROAS
What exactly does incrementality mean in this context, what are incrementality tests and how do they show the true iROAS of advertising expenditure?
In incrementality tests, users are divided into a test group and a control group, and the performance of the two groups is then compared. The test group sees the advertising campaign, the control group does not. This measurement offers advantages compared to SKAdNetwork such as real-time data or additional information, such as retention data. The method itself comes from science. For instance, it is used to test the effectiveness of a vaccine. In other words, incrementality tests are rule-based attribution models that offer realistic results.
Gatekeepers in mobile marketing
Google has taken the path of customization, has adapted its advertising SDK and advises app developers to use dialog pop-ups. Is this adjustment inevitable? And doesn't it show too clearly how powerful Apple (and theoretically also Google) are as gatekeepers in mobile marketing?
Every app developer who runs advertisements or shares data with third party providers should show the dialog pop-up, otherwise they risk being kicked out of the App Store. On the one hand, Apple has created a fantastic market with ingenious new business models. Meanwhile, this market, with its many diverse players, is so big that Apple cannot plan such serious changes without involving the industry players. Apple is clearly using its market power here and setting the rules. Given the size of the market, this is not acceptable, and there is no alternative for the iOS app developer.
Proactively tackling transparency
There is a general trend towards data protection-compliant solutions, including cookieless tracking via Chrome (Google has just presented an update on alternative solutions such as Federated Learning Cohorts [FLoC]). Should app developers and marketers therefore rely more proactively on transparency in tracking regardless of requirements? Or simply react when the time comes?
Looking at the development of privacy, one can clearly say: The future of digital advertising will not be based on an ID that is passed back and forth on the web. Apple is pushing ahead here; all browser manufacturers, except Chrome, filter cross-site tracking and Chrome wants to aggregate the user data. So, in the future we will not need any pop-ups for advertising consent, because the basis for advertising will be aggregated data, and these do not need user consent. The technologies for this are in development as the industry already needs solutions for Apple’s iOS changes. Perhaps these will lay the foundations for web technologies. As a marketer, it’s important to monitor these developments and be at the forefront of data-protection compliant solutions. That will be the future.
The specifics of opting in or out
Do you think that users will read through the specific details for tracking apps that Apple is requesting before they decide for or against opting in to tracking?
The pop-up itself is relatively short and there are intensive tests being done on how to increase the opt-in rates. In the medium term, however, I expect the opt-in rates to decrease because users do not understand the added value of tracking and will associate it negatively. Additionally, you can deactivate the pop-up centrally for all apps so that it no longer shows up and ensures that you, as a user, cannot be tracked. Hardly anyone expects opt-in rates of more than 20 percent.
Providing user consent
Would you always give your consent to tracking apps in the App Store? Are there certain hurdles?
Yes, I personally would give my approval, especially for smaller apps, because I know how much added value I can add to the app that I use for free. Where I wouldn't do it would be on Facebook or Google, as they already have a dominant market position and shouldn't benefit from it any further. The good thing about Apple's Privacy Framework is that, compared to the GDPR, Facebook and Google in particular are “caught” because their data sovereignty is significantly higher than that of independent providers.
[End]
Get the latest insights on the iOS14 changes
We have launched a ID or No ID dashboard to help mobile marketing players with their future planning in a post-IDFA world. This interactive site provides daily updates for the programmatic advertising industry, which will be further impacted by Apple’s privacy-centric iOS 14.5. Be sure to add it to your bookmarks so don't miss a day.
Receive Remerge's content conveniently in your inbox