Adjust x Snapchat: Unlocking insights on SKAN 4’s path to performance

1. We’ve been talking about SKAN 4 and its features for almost a year now, what would you really like to highlight for app marketers at this stage?

   The first thing to tackle is the change to conversion values. SKAN 4 comes with fine values (possible in Postback 1), referred to within the SKAN 3 framework as the 64 CVs or 6 bits, which are mapped to revenue events and sessions. SKAN 4 also introduces coarse values (possible in all three postbacks), which can be mapped to three values: low, medium, and high. Long story short, for the fine values, we recommend copying your SKAN 3 64 CVs mapping over one for one.

   For coarse values in Postback 1, we recommend taking your fine values and splitting them into two groups, with half assigned to the medium value and half assigned to the high value. The low value should be reserved for the session, to ensure a signal is received. This is the same for Postback 2 and Postback 3, as this enables some retention measurement, while the medium and high values in these two should be leveraged for high level insights. Keep in mind that any changes you make will result in there will be an overlap between the old and new mapping.

   For Postback 1 this will be up to 4 days because of how long it takes to receive a postback. In the worst case, for Postback 3, the overlap can take up to 34 days to be resolved. You really have to be strategic about whether changes are needed and when they can happen.

2. Among the top questions at the moment is the topic of transitioning from SKAN 3 to SKAN 4. You already talked about conversion values, what are some other important things to keep in mind when making the switch?
   Sure, I would first just add that while conversion values have changed a bit, they are still the one and only way that we receive any attribution data from SKAN, and remember that for the time being, you can run SKAN 3 and SKAN 4 campaigns simultaneously, meaning you’ve got a lot of capacity to test out the new functionality.

   What I would also really like to highlight is the importance of understanding the difference between SKAN 3’s CV window and SKAN 4’s lockWindow. The former determines the period in which to update CVs post-install causing the timer to restart after each action, while the latter locks the measurement window so that the postback will not be updated after the time it is applied.

   Another piece of advice I like to give marketers to step outside of the framework of comparing deterministic journeys to SKAN postbacks. Think about ‘time’ for SKAN as postbacks rather than days. With a deterministic journey, for example, you might say, “On day 1 a user completed 10 levels” for a gaming app. With SKAN postbacks, we would say “In postback number 1, we received 10 completed levels”. The reason to shift this mindset is because when working with SKAN 4 reporting, it can be tricky to understand when an event took place because the measurement windows are so long.

3. What about CV nulls and nones?
   Oh, of course. Nulls are what we receive when the privacy threshold is not met on SKAN 3, so the conversion value is masked and we don’t understand how the user is interacting with the app because the null is covering that information. With SKAN 4, there are also conversion values sent with a value of ‘none’ and this is when a user is part of a campaign that actually does meet the privacy threshold for a coarse conversion value, but the conversion event did not occur.

4. How are Adjust and Snapchat working together to help clients succeed on SKAN 4?
   As Ashwin Pradhan shares in the webinar, Snapchat has fully completed its SKAN 4 integration, which includes being integrated with Adjust’s Conversion Hub. For apps they’re working with and testing with on SKAN 4, they are receiving postbacks that contain all the SKAN 4 elements, like fine values and coarse values for all three postbacks. So if you’re an Adjust client running campaigns on Snapchat,  you can fully unpack and analyze your SKAN 4 postbacks within Adjust’s Conversion Hub and attribute the installs and engagements accurately.

   One of the most interesting points from their testing so far is that post-install event volume is much higher on SKAN 4, which is the result of the ability to combine fine and coarse conversion value data from the first postback. They’ve also put out high recommendations for the number of installs needed each day to meet the privacy threshold to increase the chances of getting at least the coarse value in the first postback. Based on SKAN 4 testing, however, their testing so far indicates that this recommendation can be lowered without an increase in nulls. This is definitely optimistic for advertisers.

5. Without getting into the technical details too much, how does Adjust’s Datascape help to simplify reporting on all of this?
   We’re building a lot of widgets for the SKAN 4 dashboard, including a monitor for the CV=0 bug so you can see when you’re receiving more than you should be. We’re also looking into how the different postbacks differentiate in terms of the coarse values, which is also possible for fine values in postback 1, to keep track of your user base re. When most interactions and installs are being delivered. Thinking of crowd anonymity, we show exactly which of your partners are delivering the best tiers and how you should optimize your campaigns and budgets based on that.

Any Adjust and Snapchat advertisers who want to test SKAN 4, please get in touch with your contact person from either side, we’ll be super happy to help you make the most of our integration. We want to encourage you to explore all of the resources available to you to stay on top of the developments and how you can make the most of them. As always, we’re here to answer any and all questions.