Crowd anonymity

What is Crowd anonymity?

Crowd anonymity is Apple's privacy mechanism within the SKAdNetwork (SKAN) framework that aggregates individual user data into groups before sharing it with advertisers, making it impossible to identify any single user within a dataset. It operates through a tiered system in SKAN 4.0, where the level of detail in conversion postbacks depends on how many installs a campaign accumulates within a given source. This ensures advertisers receive actionable campaign performance data while preserving the privacy of every individual user.

How it works

Crowd anonymity works by requiring a minimum number of installs before SKAdNetwork releases detailed attribution data. Apple assigns each campaign to one of four tiers based on install volume, and each tier controls what conversion data is included in the postback sent to the ad network.

Tier 0

Tier 0 represents the lowest install count, falling below Apple's minimum privacy threshold. Because the group is too small to ensure anonymity, postbacks at this tier include only a 2-digit source identifier, no conversion values in any postback, and no second or third postbacks are sent.

Tier 1

Tier 1 meets the minimum privacy threshold but still sits at the lower end of crowd anonymity. Postbacks include a 2-digit source identifier and a coarse-grained conversion value. No second or third postbacks are generated.

Tier 2

Tier 2 represents a higher level of crowd anonymity. Advertisers receive a 2-, 3-, or 4-digit source identifier in the first postback and a 2-digit source identifier in the second and third postbacks. The first postback includes a fine-grained conversion value, while postbacks 2 and 3 carry coarse-grained conversion values.

Tier 3

Tier 3 is the highest crowd anonymity level and unlocks the most granular data. It mirrors Tier 2 in conversion value structure but additionally includes the source app ID in the first postback, enabling advertisers running in-app ads to identify which publisher app drove the install.

What the tier controls

The crowd anonymity tier determines four specific postback elements: whether the postback is sent at all, the type of conversion value included (fine-grained, coarse-grained, or none), how many digits of the source identifier are disclosed, and whether the source app ID is revealed.

Why it matters

Crowd anonymity directly determines the quality of attribution data available to advertisers on iOS. Campaigns that do not generate sufficient install volume are locked into lower tiers, receiving little to no conversion detail. This creates a meaningful incentive for advertisers to concentrate budget on higher-performing campaigns and publishers capable of driving install volumes above Apple's privacy thresholds.

For mobile measurement partners like Airbridge, correctly interpreting the tier assigned to each postback is essential for accurate campaign reporting. Since the source identifier length and conversion value type vary by tier, measurement logic must account for these differences to avoid misattribution or data gaps.

Marketers operating under SKAN 4.0 need to design their conversion value schemas with crowd anonymity tiers in mind. A fine-grained conversion model is only useful if campaigns are consistently reaching Tier 2 or Tier 3. For campaigns that frequently land in Tier 0 or Tier 1, coarse-grained conversion values become the primary signal, and measurement strategy should adapt accordingly.

How to measure campaign performance under crowd anonymity tiers

Understanding which tier your campaigns fall into is the first step toward interpreting SKAdNetwork data accurately.

1. Monitor install volume per source identifier. Crowd anonymity tiers are assigned based on install counts aggregated at the campaign and source level. Track installs per ad network and campaign to anticipate which tier each will reach.

2. Design conversion value schemas for multiple tiers. Because lower-tier campaigns return coarse-grained or no conversion values, define coarse conversion value mappings that still yield actionable signals. Do not rely solely on fine-grained values.

3. Use your MMP's SKAN reporting to identify tier distribution. Airbridge and other MMPs decode SKAN postbacks and can surface tier-level breakdowns, allowing you to see which campaigns operate in data-rich versus data-limited conditions.

4. Consolidate budget toward high-performing sources. Campaigns fragmented across many small ad sets are more likely to fall into lower tiers. Consolidating spend toward sources with consistently high install volume helps push campaigns into Tier 2 or Tier 3.

5. Account for postback delays. SKAN postbacks are sent after a timer window expires and may arrive with variable delays. Ensure your reporting pipeline handles late postbacks without dropping or duplicating records.

6. Supplement with aggregated probabilistic modeling. For campaigns stuck at Tier 0 or Tier 1 due to low volume, probabilistic modeling and incrementality testing can provide directional insights that SKAN data alone cannot.

Related concepts