I'd like to report an issue with the "App Referrer" source segmentation in the App Store Connect Analytics API.
Issue Summary
When retrieving impression data via the App Store Connect Analytics API using groupBy=["sourceType", "sourceInfo"], the impressions attributed to specific referring apps (e.g., Facebook, Instagram) do not match the values displayed in the App Store Connect web dashboard.
Details
The total impressions in the API and dashboard are completely consistent, with the report category: App Store Discovery and Engagement Standard_Daily.
The aggregated "App Referrer" (overall category) is also consistent.
However, the detailed segmentation data in "App Referrer" is inconsistent, with the report category: App Store Discovery and Engagement Detailed_Daily.
For example:
The web dashboard shows significantly more impressions attributed to Facebook or Instagram.
The API returns fewer impressions for the same referral source.
Furthermore, the API does not return any "unknown"/"other" referral source entries, and displays no data for presentations where the exact referring app cannot be identified.
Therefore, the sum of all sourceInfo entries returned by the API is lower than the app referral source value displayed in the web dashboard.
We have confirmed:
This is not due to time zone differences—the totals are exactly the same.
We used the same date range, region, and platform settings as the dashboard.
The difference only appears in the detailed app referral source breakdowns (e.g., "Facebook," "Instagram").
The API does not return missing presentations under any category.
Questions for Apple:
Does the App Store Connect web interface aggregate presentations from unidentified or privacy-restricted referring apps into known app names (e.g., Facebook or Instagram)?
Is it normal for the Analytics API not to expose these presentations when the exact referring app cannot be identified?
What official logic does the App Store Connect use to determine and display specific app referral sources (e.g., Facebook, Instagram)?
When the API returns fewer referral records than the dashboard, how can developers accurately reproduce the same detailed app referral breakdown data in their reports?
Are there any known limitations or privacy thresholds in the API that affect the disclosure of sourceInfo?
Other Information
Endpoint: Analytics API (Impressions grouped by sourceType and sourceInfo)
Metric: Impressions
Grouping Criteria: sourceType, sourceInfo
We want to understand how App Store Connect allocates impressions to specific third-party apps (e.g., Facebook, Instagram) in the dashboard, and why the Analytics API returns lower numbers for these referral sources.
