SKAdNetwork (SKAN) is Apple's framework for measuring app-install advertising on iOS without exposing user- or device-level data.
Definition
SKAdNetwork (SKAN) is Apple's framework for measuring app-install advertising on iOS without exposing user- or device-level data. Instead of a per-user attribution record, the ad network receives a privacy-preserving postback from Apple confirming that an install happened and carrying a limited "conversion value" that the advertiser configured to encode a small amount of post-install behaviour, delivered after a randomised time delay and subject to thresholds that suppress data when volumes are low.
SKAN exists because App Tracking Transparency (ATT) requires explicit user permission for the cross-app identifier (IDFA) that deterministic mobile attribution relied on, and most users decline. SKAN is the sanctioned alternative: aggregate, delayed, coarse, and lossy, but usable for comparing campaigns.
A successor, AdAttributionKit, extends the same model.
In context
For iGaming advertisers running app-install campaigns on iOS — sportsbook and casino apps in regulated markets — SKAN reshaped measurement. The old model of tracking each install to a source and then optimising on that user's FTD and LTV is not available for the ATT-declining majority.
Instead, the advertiser designs a conversion-value schema that packs the most decision-relevant early signal into the few available bits: typically registration, first deposit, and a coarse deposit-value tier within the measurement window.
The consequences are practical. Optimisation windows are short and delayed, so campaigns cannot be steered on same-day FTD the way web campaigns can.
Low-volume campaigns and granular breakdowns hit privacy thresholds and return null, so testing needs concentrated spend. Web funnels (web-to-app, or web-only registration then app usage) partly sidestep SKAN by keeping the conversion on the web where server-side tracking still works, which is one reason many operators push a web registration even for app-centric products.
Android's Privacy Sandbox is moving in a similar direction, so the SKAN way of thinking — aggregate, modelled, schema-designed measurement — is becoming the general case, not an iOS quirk.
Worked example
An operator launches an iOS casino app and configures a SKAN conversion-value schema encoding registration, FTD, and three deposit-value tiers within a 48-hour window. Campaign comparison now runs on modelled FTD-tier distributions per network rather than per-user LTV, and the operator keeps a parallel web registration funnel so a portion of acquisition stays fully server-side measurable.
Related terms
Frequently asked questions
Browse the full iGaming & affiliate glossary — hundreds of EN/RU terms with examples.
← Back to glossary