> ## Documentation Index
> Fetch the complete documentation index at: https://docs.smartwpplugins.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Duplicate customer finder
> Find customer accounts that look like the same person
Duplicate customer finder scans your customer accounts for people who signed up more than once: the same email, phone or address, or a similar name at a similar address. It only lists them and never changes the accounts.
**Shows in:** Powerups, Find duplicates tab
## Turn it on
1. Go to **WooCommerce → Powerups** and switch on **Duplicate customer finder**.
2. Open the **Find duplicates** tab (also linked from the Powerups list) and click **Scan customers**.
## Scanning
A progress bar shows while customer accounts are checked in small batches: "Checked X of Y customers…", then "Comparing likely matches…", then "Grouping results…". At 99% it says "Finishing…" while the scan tidies up.
* Big stores (tens of thousands of customers) scan without timeouts. If a step is still running on a slow server, the next one waits and retries by itself.
* If the connection drops at the last step, the tab still shows the results with "Scan finished."
* Starting a scan in another tab stops the first one: "Another scan started. Scan again."
* Results stay until the next scan. Click **Scan again** to refresh them.
## What counts as a match
| Match | How it's compared |
| - | - |
| **The same email** | Gmail dots and "+tags" are ignored. |
| **The same phone** | Country code and leading zeros are ignored. |
| **The same address** | "Street" = "St", postcode spaces ignored, and the names must be close too. |
| **A similar name at a similar address** | Typos, word order, names that sound alike (Catherine and Kathryn) and initials. |
Families at one address aren't flagged: when surnames match, first names must match too.
Only customer accounts are checked: not staff, and not guests.
## Results
Each group shows a match score, the reasons, and each account's email, phone, address, orders and join date, with links to the accounts.
* The **Orders** column leaves out checkout drafts and trashed orders.
* The tab shows the 100 most likely groups at a time. Mark groups **Not a duplicate** to see more.
* With very many matches, the strongest are kept and the tab says "Showing the strongest matches. There were too many to list them all."
### Not a duplicate
**Not a duplicate** hides the group for good. It covers the whole group as shown (up to 200 accounts). If the group has changed since (after a newer scan), you're asked to use the current results: "That group is no longer in the results."
If a scan step or another change is running at that moment, nothing is saved and you see "A scan step or another change is running. Try again in a moment."
You can bring dismissed matches back by deleting **Marked "Not a duplicate"** in the Data section.
## Settings
| Setting | Default | What it does |
| - | - | - |
| **Match customers with** | All four | **The same email**, **The same phone**, **The same address**, **A similar name at a similar address**. |
| **How close names must be** | Balanced | **Strict**, **Balanced** or **Loose**. Looser finds more, with more false matches. |
A scan reads every customer account in batches. On stores with many customers it can take a few minutes. It runs only when you click Scan.
## Multisite
Only this site's customer accounts are checked and shown. Accounts removed from the site drop out of the results.
## Data
| Item | What it is |
| - | - |
| **Settings** | This power-up's saved settings. |
| **Last scan results** | The groups found by the last scan. Scan again to rebuild them. |
| **Scan in progress** | Where a running scan is up to, so two scans at once don't mix. Removed when a scan finishes. |
| **Marked "Not a duplicate"** | Pairs you dismissed. Deleting them lets those matches show again on the next scan. |
| **Scan work tables** | Small database tables used while scanning. Emptied after each scan; deleting them is safe, the next scan creates them again. |
## Known limitations
* Names in non-Latin scripts are compared exactly. Sound-alike matching works for Latin names.
* At most 1,000 groups are kept, strongest first.
* If the server kills a scan step, the next step can start after 2 minutes.
* "Not a duplicate" pairs are kept in one list that grows with every dismissal. Clear it in the Data section.
* A scan abandoned while tidying up is cleared by one larger delete when the next scan starts.
This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.