---
title: How CSV duplicates are merged
description: How an import decides a row is someone you already have, and what it changes on their record.
search:
  keywords: [duplicates, dedupe, merge, same email, same phone, skipped rows, existing contacts, reimport]
related:
  - /import-and-export/import-from-a-csv-file
  - /contacts-and-leads/duplicates
  - /contacts-and-leads/delete-contacts
---

A row is a duplicate when its email or phone matches a contact you already have, or an earlier row in the same file, and the names agree. Duplicates merge into one contact instead of creating a second.

## How a match is found

- Email first, ignoring capital letters, then phone number.
- The names must agree. A row with a different first or last name stays a separate contact, so a couple sharing an email or a home phone stays as two people. A blank name, or "Unknown", agrees with any name.
- When several contacts could match, the oldest one is used.

## What a merge changes

A merge never overwrites anything you already have:

- Blank fields are filled in from the row.
- New tags are added to the existing ones.
- A different imported note is added below the existing one.
- The lead type of an existing contact stays as it is.

## The results

**Import complete** sums it up, for example "12 new contacts, 3 duplicates merged, 1 row skipped". Below it:

- **Merged into one contact** lists each person a row merged into, and whether it matched on the same email or the same phone.
- **Skipped** lists rows with no name, email or phone.

:::note
If a few existing contacts couldn't be updated, the results say so. Import the file again to retry them; nothing is created twice.
:::