At the Worldwide Developer Conference 2022 (WWDC) Apple presented a new version of the SKAdNetwork API. The fourth version of SKAdNetwork is set to release later this year. The new version is aimed at providing advertisers with more data so they can better measure and optimize the performance of their apps, while maintaining the privacy of their users simultaneously.
The SKAdNetwork 4.0 changes can be grouped into 4 main categories:
- Hierarchical source identifiers
- Hierarchical conversion values
- Multiple conversions
- SKAdNetwork attributions for ads on the web
In the video below, we present a Q&A session with our co-founder and CEO, Christopher Farm. He tells us what the SKAdNetwork 4.0 changes are, when we can expect to see them in action, what the changes mean for advertisers, and how Tenjin feels about these changes.
SKAdNetwork 4.0: what’s changing, when, and what does it mean for advertisers?
Question: What are some of the most important SKAdNetwork updates to know?
Answer: The biggest changes in my opinion are that SKAdNetwork 4.0 will now have the ability to track multiple conversions, possibly do some web attributions later in the future.
Multiple conversions are one of the bigger changes because it will start to provide advertisers more capabilities to look at things on a cohorted basis. There are some things within the multiple conversions framework that generally are not as fine-grained as before. In fact, they call it fine-grained versus coarse-grained.
So they’ll fall into 3 buckets of multiple conversions on a cohorted basis. So within the first 0-2 days you might get a fine-grained conversion value which will be a specific conversion value 0-63. But in the second two cohorted buckets, where you can see more of a lifetime value type of a perspective, it would be more of a coarse-grained version of this (high, medium or low). But it does give you more information as an advertiser as to the lifetime performance of a specific cohort whereas that didn’t really exist before when you got only one conversion value from Apple. In the long term, this is going to be better for the advertisers because now the ad networks are receiving these SKAdNetwork values, whether it is a coarse-grained or fine-grained cohorted value. They can use that information to help do optimizations as well.
Question: When can we expect to see these SKAdNetwork 4.0 changes in action?
Answer: The answer given by Apple is later this year. I also do think they will roll out each type of incremental development at different time frames this year.
Question: What do the SKAdNetwork 4.0 changes mean for advertisers?
Answer: This is a harder question to answer because we don’t know how ad networks specifically will use this new information. We also don’t know to what extent Apple will enforce the use of this information. There are still a lot of networks that use probabilistic attribution who don’t necessarily need this information to continue doing what they’re already doing. As Apple starts to enforce more ad networks to fall within this framework, they’ll probably end up using it in a lot of different ways within their own capabilities. So Meta obviously has a lot of capabilities that are different from Google or Snapchat in terms of targeting their audiences and so the relative nature in which they use some of these multiple conversions will all be different. Ultimately for advertisers, they’re just going to have to figure out which channels provide them with the best conversion value to spend ratios depending on whatever channel they’re working with, and however that channel decides to optimize those values.
So there will be a pretty integral part that MMP’s will play. They’ll have to figure out and recommend to customers what the best ways are to set up their conversion values and what the best ways are to get them to be performant on the different channels that exist.
I think we’re still in the very early stages of the SKAdNetwork lifecycle for a number of reasons. I think the primary thing that everyone is just itching to learn more from Apple is how and when all of these changes are going to be enforced. I think Apple is in a tough spot to decide how to enforce this given all of the environment we’re in on privacy, as well as how they’re looked at from controlling the App Store. They’re kind of in a tricky spot, they’re trying to solve the problem from both ends while also helping app developers do what they need to do.
I think there’s a lot of advertisers just waiting to see what happens in terms of enforcement for one, but I also see and talk to a lot of developers who quite frankly are just trying to figure out what works best for them. They mostly start off on channels where these changes are being enforced to try to figure out for themselves how the conversion values work. There’s some 10-20% budget to figure out how SKAdNetwork would work with a specific channel, and how it could be optimized. I think this share is increasing overtime but it will definitely increase much faster once, or if, Apple decides how they want the ecosystem to play out.
Question: Do you feel like it’s been a sandbox for advertisers and developers to try to figure out how to use SKAdNetwork so far? Do you think the new API changes make it more clearly defined how you can use SKAdNetwork to get the value you are looking for?
Answer: Yeah it’s definitely still a sandbox. I would say everything pre-SKAdNetwork 4.0 was even a smaller sandbox. Apple got a few ad networks to enforce the capability for SKAdNetwork. I think Apple is just iterating with SKAdNetwork 4.0 to accommodate a lot more options for these ad networks and help them out.
The real tipping point here is going to be something to do with when advertisers start to see parity between the old way of doing things with advertising ID and deterministic matching as well as probabilistic optimization methodologies.
Once SKAdNetwork has enough to get an advertiser to parity in terms of their performance and ROI, there would be no question as to why an advertiser shouldn’t switch to something like SKAdNetwork with equal parity of ROI. I think Apple is in a position where they’re trying to get to some level of parity with the current channels that exist. And once they get to a high enough state where they feel there is parity, I think advertisers will naturally prefer the more privacy-centric solution.
Question: How does Tenjin feel about SKAdNetwork 4.0 changes?
Answer: I think it’s definitely progressing well. It’s progress for ad networks and advertisers who rely on those network optimizations. But I also think there is still a lot more enforcement that is needed in order for an attribution provider, or someone like Tenjin, to provide a lot of value.
I do think there is some value in the fact that we can recommend some conversion value schemas that work better than others on specific networks. That is one area that we’re investing a lot of effort in.
I think we’re also in this sandbox phase with a lot of our customers, trying to figure out how to configure their applications with SKAdNetwork to the point where they can get to parity with their returns. The position we’re taking here is, can we build a product that gets a customer to ROI parity on a privacy-centric platform, and if so that’s going to be the de facto way to advertise.