Glossary · Business Analytics
SKAdNetwork
also: SKAN · SKAdNetwork 4 · Apple privacy attribution · SKAN 4.0
Definition
SKAdNetwork is Apple's privacy-preserving attribution framework for iOS app install campaigns. It delivers aggregated, delayed, randomized-timing postbacks with a sparse conversion-value payload, and requires roughly 25 conversions in a privacy bucket before any data is released to the advertiser.
SKAdNetwork (SKAN) replaced the IDFA-based deterministic attribution model that dominated iOS user acquisition through 2020. SKAN 4 (introduced 2022) supports three postback windows, fine and coarse conversion values, and crowd anonymity thresholds. The privacy threshold (typically about 25 conversions in a bucket) means small campaigns or fine-segmented audiences return null data; the postback timing randomization (0-24 hours) breaks deterministic time-of-day join logic. Operators have responded by combining SKAN with geo-holdout incrementality designs, MMM, and CAPI server-side conversion forwarding to reconstruct what last-click attribution can no longer deliver.
Essays on this concept
- Business Analytics
The Death of Last-Click in Mobile-App Attribution
Why SKAdNetwork 4 postback loss, IDFA opt-out rates, and the Apple privacy threshold have ended last-click attribution for mobile apps, and how incrementality testing has become the operational ground truth.
- Marketing Engineering
Multi-Touch Attribution Is Broken, A Causal Inference Approach Using Directed Acyclic Graphs
MTA models overestimate retargeting by 340% and underestimate display by 62%. The fix isn't better heuristics, it's abandoning correlational attribution entirely in favor of causal graphs.
- Marketing Engineering
Marketing Mix Modeling in the Privacy-First Era: Bayesian Structural Time Series Without User-Level Data
Cookies are dying. Deterministic attribution is shrinking. The irony: the measurement approach from the 1960s, Marketing Mix Modeling, is making a comeback, now powered by Bayesian inference that would have been computationally impossible when it was first invented.
- Marketing Engineering
Unified Measurement Architecture: Connecting MMM, MTA, and Experimentation Into a Single Source of Truth
MMM says Facebook works. MTA says Google works. The incrementality test says neither works as well as you thought. Three measurement systems, three different answers, here's how to reconcile them into one coherent picture.
- Business Analytics
Mobile App SDK Overhead vs. Telemetry Value
Most mobile apps over-instrument. The cost shows up in binary size, cold start, battery, and privacy permissions. This essay maps the SDK trade-off honestly, with the question of what to drop and what to keep.
Authoritative references