What is Deduplication?

Deduplication is the work of finding records in a CRM that describe the same person or company, merging them into one, and setting rules so new duplicates get caught before they’re created.

At a glance

• A duplicate is two or more records for one real person or company: two contacts for the same buyer, or two accounts for the same business.
• HubSpot deduplicates contacts by email address and companies by domain name automatically, according to its own help docs.
• Salesforce uses matching rules (which fields to compare, exactly or loosely) and duplicate rules (whether to warn the user or block the save).
• It has two halves: cleaning up what’s already there, and preventing new duplicates at the point of entry.
• The hard part isn’t finding duplicates. It’s deciding which value survives the merge.

How deduplication works

Every deduplication setup starts with a definition of “the same”. For people, most teams match on email, because two different people can share a first and last name. For companies, the usual key is the website domain, with a looser match on company name to catch “Acme Inc” and “Acme Incorporated”.

The rules differ by CRM. HubSpot updates the existing contact when a form or import brings in an email it already has. Its docs also note two gaps: an import without an email or another unique ID creates a new record for every row, and companies created through the API, including third-party sync apps, aren’t deduplicated by domain at all. In Salesforce, a duplicate rule can alert the rep or block the save, and Duplicate Jobs, which scan the whole database, are only included in the Performance and Unlimited editions.

Then comes the merge. When two records become one, someone has to decide, field by field, which value is the source of truth: the newer email, the original lead source, the right owner. Activity history should carry over to the surviving record so nobody loses the thread.

Example (illustrative): a buyer fills in a demo form as [email protected], and a rep later adds her by hand as [email protected] from a list. Now there are two contacts, two owners and two sets of emails. A rule that flags contacts sharing a domain, first name and last name catches it, and the merge keeps the form record’s lead source and the rep’s notes.

Common deduplication mistakes

• Treating it as a one-time cleanup. If imports, forms and integrations keep creating records with no rules in front of them, the duplicates come back.
• Matching on email only. People change jobs and use personal addresses, so exact email matching misses a lot of real duplicates.
• Blocking too hard. A rule that blocks every possible match stops reps from saving real records, and they find a way around it.
• Merging without merge rules. If nobody decides which value wins, a stale record can overwrite a qualified lead’s status or owner.
• Forgetting leads versus contacts. In Salesforce, a new lead can duplicate an existing contact, and that needs its own rule.

How deduplication shows up in a revenue system

Duplicates do the most damage in outbound. One person on two records can get the same sequence twice from two reps. At Datatruck, building the outbound motion around one embedded AE included setting up HubSpot from scratch, and a new CRM is the cheapest moment to set matching rules. Read how the Datatruck sales motion was built.

Terms that sit next to deduplication

Deduplication is one part of data hygiene, and it matters most right after data enrichment or an import, when new records arrive in bulk. Clean field mapping keeps an integration from writing a second record instead of updating the first. Duplicates break lead routing, because the same buyer can land with two reps, and they put the same person into a sequence twice. All of it lives in your CRM.