Find and Process Duplicate Records

Last published at: September 3rd, 2024

This article is also available as a how-to video. Watch now.

 

Cleaning up your data consists of two steps: first you find out which records are duplicates of each other, and then you consolidate them into one record.

1.  Finding duplicate records

There are several ways to find duplicate records in your org. They all make use of Scenarios, where you have defined when two records can be considered potential duplicates of each other, based on their field values. You can define multiple scenarios per object, to search for duplicates in different ways.

Manual insert or update Using DC Entry and DC Live, you will be alerted if a record already exists when creating, editing or opening a record. The duplicate records are shown so you can decide how to proceed.
Run a DC Job Run a Duplicate Check batch job to search in your org for duplicate records. You can search within one object, between different objects with Cross Object, or on a subset of records by applying filters to the job. You can schedule a job to run repeatedly at a set day and time.   
The found duplicate records can be reviewed via the DC Job Overview.
Automatically check upon insert Set Duplicate Check to run a duplicate search when one or more records are inserted or updated in your org via Web-to-Lead, Unique Import, or API Insert/Update. A DC Job is run automatically, and the found duplicate records can be reviewed via the DC Job Overview.

 

2. Processing duplicate records

Once you have found duplicate records in your org, you can process them to clean up your data. You can merge them together into one record; if you found duplicates between two objects (e.g. a Lead that already exists as a Contact) the merging is called “converting”. Or you can decide to mark the duplicate records as being similar, but keep both; at a next search they won't be brought up as potential duplicates anymore.  
If you don't want users to merge records themselves, you can have them mark records to be merged, and then let someone else perform the merge. 

Upon merge, only the Master record is kept; the losing records are deleted. Related records will be reparented to the Master record.

Manual merge or convert Manually merge or convert duplicate records to decide for each duplicate set which data from which record should be kept. You can add custom data as well.
Automatically merge or convert job results Merge or convert duplicates automatically after a job ran, with or without reviewing them first. Merging is done based on pre-set Merge Rules. Records where it's not clear whether they are duplicates can still be processed manually.
Directly merge or convert upon insert Merge or convert duplicates directly upon insert via Web-to-Lead, Unique Import, or API Insert/Update. Merging is done based on pre-set Merge Rules. Records where it's not clear whether they are duplicates can still be processed manually.
Discard records When reviewing duplicate records, you can decide to discard false positives, and not show them again as potential duplicates in future searches. If you want to merge them at a later point after all, you can undo the discard.