Introduction to conversion values: Adjust’s SKAdNetwork solutions

Success on Apple’s SKAdNetwork (SKAN) comes down to understanding its conversion value mechanism and how to develop a strategy that makes the most of it. A conversion value is essentially the way that app developers are able to grade acquired users (based on their post-install events) in order to measure and optimize for future campaigns. In short, a conversion value is a representation of an action that a user takes in-app. All data received from SKAdNetwork, however, is obfuscated and shared in the aggregate, meaning no granular data is accessible at the user level.

SKAdNetwork makes space for six bits of downstream metrics, a number between 0 and 63 (or between 000000 and 111111 in binary), with an initial 24 hour timer. This conversion value is configured by developers and represents the events, chain of events, or characteristics that an app/developer determines as an indication of the actual value of a user themselves. By switching the six bits on or off, there’s the possibility of 64 total conversion values that can be configured, and it’s up to the developer to determine how they use these bits, which combinations make sense, and what they want to measure.

Every time the conversion value is updated, the timer window is extended by an additional 24 hours. Once this conversion value-window expires, a second timer for attribution starts counting down. Within this window, which can be up to 24-hours, SKAdNetwork randomly returns the attribution data.  The data, including the conversion value, is received, packaged up by Apple in the SKAdNetwork postback, and pinged back to the network. Since the rollout of iOS 15, there’s also the option for developers to add an additional endpoint to the configuration of their app to also receive a copy.

Along with the conversion value, the postback includes anonymized level information on whether an install or reinstall occurred, whether it’s impression or click based, the source app ID (which is usually hidden by networks) and the campaign ID, which is used to link the conversion value to the metrics of a campaign like impressions, clicks, and costs. This is received regardless of whether a user has opted in to ATT or not as it does not include the device ID, and because of the random timer and anonymization, it’s impossible to link the data to a specific device.

This is why it’s crucial to develop a highly optimized conversion value strategy.

Let’s have a look at the solutions Adjust offers clients:

This solution enables clients to measure up to six specific in-app events in parallel to each other. It allows the measurement of which of these six events have been triggered or not triggered, in any variation or combination of the events.

With six bits of information available, this approach treats each bit as a switch, with 0 denoting an off switch and 1 signaling that the switch is on. You can tell if an event happened when the corresponding bit is activated.

When a user completes one of the events that trigger a conversion value update, the Adjust SDK sends this on to Apple. Depending on which switches are flipped, you can tell which of the six events the user completed during the conversion value window.

In order for a client to inform their app which events are the ones that should trigger a conversion value update, they can pick the events they want to track as SKAdNetwork conversions directly from the Adjust dashboard. When we receive the data we unpack it according to the relevant event metrics, which allow marketers to see the performance of each event per campaign in the reporting, or via raw data. If you have events that you’re already tracking and would like to continue tracking them via SKAdNetwork, it’s a simple process to set up.

• Just navigate to Settings > iOS 14 Support > SKAdNetwork > Manage SKAdNetwork Conversion Values.
• Here you can toggle on the option to track conversion values and map your six desired events to each of the six possible bits.

However, there are use-cases where six events are not sufficient or questions where yes/no answers don’t give us the solutions we need. In these cases, Adjust supports using the full 63 usable bits for different conversion value scenarios. With the Advanced Conversion Value Schema, you can track event ranges or value scales that encompass any measurable behavior, including:

• Events
• Sessions
• In-app revenue
• Ad revenue

Using defined schemas, you can record a count of different in-app events, so you could trigger a conversion value update on the user’s second purchase, fifth login, or third ad impression. You can also trigger value updates based on ranges, so you could set a range of $5-10 of purchase revenue and update the conversion value when the count reaches this range.