Mismatch in Analytics data between App Store Connect UI and Analytics API

We are observing a discrepancy in crash data between the App Store Connect UI and the App Analytics APIs for our iOS application.

For example:

On March 10, App Store Connect shows approximately 2,443 crashes (average ~2K/day). However, in our analytics database (fetched via App Analytics APIs), we are seeing 10,000+ crashes for the same date.

APIs Used (IDs masked):

/v1/apps/{app_id}/analyticsReportRequests /v1/analyticsReportRequests/{request_id}/reports?filter[category]=APP_USAGE /v1/analyticsReports/{report_id}/instances?filter[processingDate]=YYYY-MM-DD&filter[granularity]=DAILY /v1/analyticsReportInstances/{instance_id}/segments

Clarifications Required:

  1. What could be the reason for the significant difference in crash counts between App Store Connect UI and Analytics API

  2. Are there differences in how crashes are counted in: UI metrics vs raw analytics reports? Deduplication, session-level vs event-level aggregation?

  3. Is there any known aggregation rule, or filtering applied in the UI that is not reflected in the API data?

  4. Which source should be considered the accurate/authoritative count for reporting purposes?

Any clarification or guidance on this discrepancy would be greatly appreciated.

Mismatch in Analytics data between App Store Connect UI and Analytics API
 
 
Q