What is SKAdNetwork?
SKAdNetwork (SKAN) is Apple's privacy-centric attribution framework introduced with iOS 14. It allows advertisers to measure campaign effectiveness without tracking individual users, by providing aggregated, anonymized conversion data.
SKAN is now the primary method for attributing iOS app installs since the introduction of App Tracking Transparency (ATT), which requires users to opt-in to tracking.
How SKAdNetwork Works
- Ad Display: An ad is shown to a user within an app
- Click/View: User engages with the ad
- Install: User downloads and opens the advertised app
- Conversion Value: App sets a 6-bit value (0-63) representing user value
- Timer Expires: After a privacy-preserving delay
- Postback Sent: Apple sends anonymous data to the ad network
Conversion Values
SKAN uses a 6-bit conversion value (0-63) to communicate post-install behavior. Advertisers must carefully design their conversion value schema to maximize insights within this limitation.
- Revenue ranges: Map user spending to different values
- Engagement tiers: Track retention or engagement levels
- Event completion: Record which key events users complete
SKAN 4.0 Key Features
Up to 3 postbacks at different time windows
Coarse (low/medium/high) and fine values
Track installs from Safari web campaigns
Data granularity based on campaign scale
SKAN Limitations
- Delayed data: 24-48+ hour reporting delays
- Limited signals: Only 64 possible conversion values (6-bit)
- No user-level data: Cannot tie installs to specific users
- Timer resets: Conversion window extends with each update
- Null values: Some postbacks contain no conversion data
Best Practices
- Design conversion value schemas that capture your most valuable signals
- Use predictive models to supplement limited SKAN data
- Run larger campaigns to unlock more granular data (crowd anonymity)
- Combine SKAN data with other signals for fuller picture