Engineers are currently investigating the issue.
Summary:
Segment creation or targeting via CSV file upload fails or throws a processing error when the file contains data formatting anomalies—such as duplicate rows, truncated leading zeros, unrecognized IDs, or relational IDs that match multiple users (such as Course IDs).
Expected behavior:
The CSV upload wizard should cleanly parse files, automatically deduplicate or skip non-breaking discrepancies, and successfully process valid entries while surfacing a clear, itemized error log for any rejected rows.
Workaround:
- For Duplicates and Formatting: Manually sanitize the CSV file using spreadsheet software to remove duplicate entries. Ensure the ID column is explicitly formatted as "Plain Text" before saving to preserve necessary leading zeros.
- For Relational Fields (1:Many Mismatches): For metadata fields that map to multiple users (such as Course IDs), manually input the values directly into the segment builder interface using separate "OR" condition blocks instead of a bulk CSV upload.
- For Debugging: Open your browser's Developer Console (
Right-Click > Inspect > Console) prior to clicking upload. The console log will frequently capture the specific row value causing the ingestion backend to reject the file.
Steps to reproduce:
- Navigate to the custom segment creation screen.
- Select a target metadata attribute field and choose the CSV contains rule dropdown.
- Upload a CSV file that contains duplicate values, truncated strings missing leading zeros, pre-provisioned IDs that haven't logged in yet, or relational IDs matching multiple users.
