Duplicate contacts in a CSV import: reconcile rows with people

Define the decision

A source file describes rows, while an outreach database describes contacts and memberships. One person can appear in several exports, and the same email can be repeated with different company text. Decide which information is trustworthy before treating the latest row as an update instruction.

Work through the procedure

1. Preserve the original file and create a comparison copy. Normalize surrounding whitespace and inspect repeated addresses without discarding the original values.
2. Separate duplicates within the file from addresses already present in the workspace. These categories answer different questions about the result.
3. Review conflicting fields manually, especially names, company and status. Do not assume reimporting overwrites existing records or clears prior suppression.
4. Compare import outcome counts with representative contacts and list membership. Record why new records differ from source rows.

Worked example

The following is a synthetic example for this procedure, not a customer result or performance benchmark.

Source rows: 100
Repeated addresses within file: 10 extra rows
Distinct remaining addresses: 90
Already present: 20
Potential new records before other validation: 70
List membership may include existing contacts too

Read the result

The example's 70 is a ceiling before invalid or excluded records are considered. It is not an expected result for every import configuration. If the job reports fewer records, inspect its categories and a sample of rejected rows. Reuploading without understanding the outcome can make the audit harder.

Check before moving on

1. Verify that a repeated address with a changed name still refers to the same person.
2. Check whether existing records joined the intended list even when they were not newly created.
3. Keep suppression and active conversation history intact during cleanup.

Limits and next action

Zintara imports have separate job processing and contact creation stages. A completed job does not mean every row became a new contact. Use the job result and actual records together; do not delete contacts merely to force the imported count to match a spreadsheet.

Source: Zintara product and contact workflow context