To measure user quality and optimize campaigns on iOS, developers today are building strategies around the SKAdNetwork conversion value. While it's the main way to get accurate insight into user quality, there are many challenges that come with it:
- It’s universal - all of your marketing channels will receive the postbacks based on the same conversion value schema
- Only one solution or platform can update the conversion value - whether it’s your own in-house solution or a third party solution
- Changing your strategy is tricky - if you decide to change your conversion values, it's critical to understand that it will harm both measurement and optimization across all your campaigns
Adoption picks up pace
Since Apple introduced SKAN 2.0 more than a year ago, most developers already have a strategy in place for measuring it today. By now, the majority of them are using conversion values to do so. In fact, according to ironSource data, adoption of conversion value (CV) has picked up since Apple fixed a bug in November 2020, and accelerated with the push of iOS 14.6.
Today, there are multiple strategies for mapping conversion values, but two dominate: the most common is mapping in-app events to bits, followed by revenue measurement. Let’s dive into each.
Strategy #1: In-app events to 6 bits
A bit, or binary digit, is a basic unit of information that can be one of two possible values - typically 1 or 0. Using bits, SKAdNetwork allows you to measure whether specific in-app events happened (at least once) or not, without any priority or sequence.
As a developer, you’re limited to measuring up to 6 events. That’s because SKAN allows 64 conversion values - with 6 bits, in which each one measures whether an event occurred or not, then there are exactly 64 possible combinations.
Here’s an example of this in action - note that this data would continue all the way through to CV 64. CV 10, for instance, would indicate that the user reached level 20 and subscribed.
Pros and cons
The 6 bit strategy lets you understand whether specific events, that you’ve defined as good signals of user value, occurred or not. In addition, theoretically, you can use a 6-bit strategy to optimize differently on several networks: for example, on Facebook you could optimize towards a specific IAP event, and on Google you could optimize towards a specific level completion, like the user reaching level 20.
The main drawback is that it’s not effective for games that aren’t able to correlate events in the first 24 hours post-install to user value - which, beyond hyper-casual, include many games.
Revenue measurement
The second most common strategy for conversion value management gives a revenue-based value to each conversion value. It's important to note that traditionally, IAP revenue has been the only form of revenue available. There are a few ways to do this:
Counting dollars or cents
With this approach, each conversion value represents a specific amount of revenue that the user has generated. For example, conversion value 1 could be $0.99, conversion value 2 could be $1.50 - following this pattern all the way up to conversion value 63.
The revenue range of your 63 conversion values will depend on your internal benchmarks - for example, if your data shows that your users generate between $0.99 to $10.99 within their first 24-48 hours, you’d logically start your conversion value 1 at $0.99.
Buckets
Alternatively, some developers split their conversion values into buckets. For example, CV 1 could equal everything between $0.99 to $2.99, CV 2 equals everything between $2.99 and $5.99, and so on. The range of each bucket can differ - for instance, CV 3 could be everything between $10.99 and $25.99. It’s up to you to test it and determine which range per bucket makes the most sense for your game.
The idea is that once you get the postbacks, you can then work out the revenue average from each bucket. This in turn provides you with a greater range of users’ value. The benefit of this is that you then have the data needed to optimize your UA bidding strategy towards the users with the highest CVs.
By contrast, if each conversion value is assigned a specific number, like in the first approach, it's possible that most of your traffic will be on the lower end of your conversion value map and just a small amount will reach the top values. As a result, even though you want more of the top users, you’ll find it very difficult to efficiently optimize towards these users because you lack the necessary data.
Pros and cons
The beauty of this strategy is that there’s no better proxy for paying users than paying users themselves. This approach fits well for games that can properly monetize users within the first 24 hours and have a statistically significant part of their ARPU curve occur on D0.
However, because this measurement strategy traditionally excludes ad revenue, ad-based developers such as those making hyper-casual games have been left without an effective revenue-based measurement solution - until now.
How to measure user value based on ad revenue
To correlate conversion values to ad revenue, ad-based game developers need a solution by an MMP or mediation platform. For example, ironSource’s conversion value manager, which is available to ironSource mediation partners, provides ad revenue insights that enable developers to continue measuring D0 ARPU and optimize towards D0 ROAS across the board.
Start managing your conversion value strategy with ironSource’s CV Manager solution inside our iOS toolkit. Learn more here
