How to Merge Duplicate Records in Salesforce (and Undo a Bad Merge)

Where Salesforce duplicates come from

Most orgs don’t have a duplicate problem. They have three duplicate sources that keep running.

• Imports. A trade show list or vendor file loads without checking what’s already in Salesforce, and every existing contact gets a twin.
• Manual entry. A rep creates a new lead instead of searching first, often with a nickname, a typo, or a different email.
• Forms and integrations. Every web form, marketing automation sync, and connected app is another door records can come through.

This is why one-time cleanups don’t last. If you merge 5,000 duplicates and leave all three doors open, they come back. A real fix covers both cleanup and prevention.

How to merge duplicates natively in Salesforce

Salesforce includes basic tools for merging duplicates. For a handful of records, they work.

1. Set up matching rules (what counts as a duplicate) and duplicate rules (what happens when one is found).
2. Open a record that’s flagged as a potential duplicate, or open Duplicate Record Sets from the App Launcher.
3. Select Compare and Merge, choose the master record, and pick which field values to keep.
4. Confirm the merge.

Where native merging falls short

• It’s one set at a time. Native merge handles up to three records per merge, which doesn’t scale to thousands of duplicates.
• Matching is limited. Rules catch close matches but can miss nicknames, abbreviations, and formatting differences like Pat vs. Patricia or Street vs. St.
• There’s no business-logic master selection. Someone has to pick the winning record by hand every time.
• You can’t undo it. Salesforce warns you before merging because once records are merged, they can’t be unmerged.

That last point is why many admins avoid merging at all and let duplicates pile up.

Step 1: Find duplicates with filters

In Cloudingo, finding duplicates has never been easier. A filter is a saved report of records that meet your matching criteria (including recommended filters to help get you started). You choose which fields to compare, combine them with “and” or “or” logic, and set a match style for each field, from exact to fuzzy.

A practical starting filter for contacts:

• First name: synonym match (catches Pat and Patricia, Bill and William)
• Last name: exact match
• Email: exact match

You don’t have to build from scratch. Start from a recommended filter, or clone one you already trust and change a single field. Swapping first-name matching from exact to synonym takes seconds.

Tip: Run a new filter and review the results before you merge anything. If it’s catching records that aren’t true duplicates, tighten it. If it’s missing obvious ones, loosen one field at a time.

Step 2: Choose the master record and set merge rules

The master record is the one that survives the merge. Choosing it well is what keeps a merge safe.

In Cloudingo, you set the rule once, and it’s applied to every merge:

• Most complete record: the one with the most populated fields
• Most recent activity: the one your team is actually working
• Custom priority: your own business logic, for example, Customer beats Prospect beats New Lead

You can also control individual fields. For example, instead of keeping one record’s picklist value and dropping the other, you can combine the values from both records.

Before committing, preview the merge side by side to see exactly what the surviving record will look like. This is the step that protects your reports and forecasts. If the wrong record wins, the wrong owner, stage, or source carries forward.

Step 3: Automate merging and prevent new duplicates

Once a filter and its merge rules are producing results you trust, stop running them by hand.

• Scheduled merges: run the filter automatically on a schedule.
• Real-time filters: merge new duplicates almost as soon as they’re created.
• Manual review: for sensitive objects, review duplicates side by side and merge them one at a time, either in Cloudingo or inside Salesforce with Cloudingo’s Lightning Web Component.

Prevention closes the import door. With Cloudingo’s Import Wizard, you drag in a CSV and it’s checked against existing Salesforce records before anything loads. Matching records are updated and only new, unique records are created.

To see whether it’s working, the Data Quality Dashboard shows where your data stands and gives guided steps to fix problems it finds.

How to undo a bad merge in Salesforce

Natively, you can’t. Once Salesforce merges records, there’s no unmerge. Your options are rebuilding the lost record by hand or restoring from a backup.

With Cloudingo Professional or Enterprise, you can keep a record of each merge so you can reverse it:

1. Open the Undo/Restore panel and find the merge.
2. Click Restore.
3. The records split back apart, and related opportunities, tasks, and notes are re-parented to the records they belonged to before the merge.

Knowing you can reverse a merge changes how you work. You can run larger cleanups and automate more, because a mistake is recoverable instead of permanent.