You can recognize a messy contact list without knowing which records are safe to combine. Two matching names might be one person using different email addresses. They might also be two people at the same company. The decision needs evidence before it needs a merge button.

TLDR: Review duplicate candidates against an identity rule, preserve the relevant record details, approve the property and association choices, then validate the merged result. HubSpot documents that merges cannot be undone. Treat prevention as a separate task, because merging existing records does not fix the process creating them. HubSpot merge documentation

Define duplicate contacts before cleaning your HubSpot CRM

We recommend writing an identity rule that your sales and operations teams can apply consistently. A name match alone is insufficient. Review each contact record: email addresses, company name, role, activity history and any reliable identifier from your source system. Your aim is accurate identity matching, not simply fewer rows in the database.

Do not use a shared mailbox as proof that two individuals are the same person. A contact called “Purchasing” can represent a team rather than a person. Similarly, an employee changing companies creates a history question that a simple domain match cannot answer.

HubSpot deduplicates contacts by email in supported creation paths. A manual attempt to create another contact with the same email address is blocked; an import using a matching email generally updates the existing contact record. Imports can use Record ID or a custom unique-value property to identify existing records; importing without an appropriate identifier can create new records. Confirm the rule for the specific entry path you use. HubSpot deduplication documentation

If duplicates come from several integrations, make identity ownership part of your HubSpot operations review. Each system should have a documented matching rule and a named person responsible for exceptions.

Find and manage duplicates in HubSpot contacts

As checked on August 31, 2026, the duplicates manager requires an eligible Professional or Enterprise subscription and the required data-quality and record-edit permissions. To review candidates, navigate to CRM > Contacts > Actions > Manage duplicates, then select the matching rule. The tool compares fields including names, email, phone and company name. These are potential matches for review. HubSpot duplicates manager requirements and navigation

If you already identified two records through search or an import review, open both and compare the evidence before using the merge action. For manual deduplication, we recommend keeping a short decision log with the original IDs, the suspected cause, the owner and the next action. Count how many duplicates are confirmed separately from candidates still awaiting investigation.

Review duplicate records before deciding to merge

The following examples are synthetic. They illustrate decisions, not observed Selworthy records.

Candidate pairEvidence availableRecommended decision
Jordan Lee, work and personal emailThe person confirmed both addresses; matching account historyReview for merge after checking consent and associations.
Alex Morgan, two company domainsSame name, different employers, no confirmationHold for identity review.
Two contacts using purchasing@example.comSeparate named buyers sharing one mailboxDo not merge merely because the mailbox matches.

Record the decision and reason even when you reject a merge. Otherwise, the same ambiguous pair can return to the queue and consume another review cycle.

Set merge rules before you merge duplicate contacts

Before an authorized user acts, we recommend these checks:

  • Identity is confirmed. Keep the supporting record or reviewer note with the request.
  • Important values are preserved. Export relevant properties and record the activity and association details you will need to verify. An export is evidence, not a promise of full restoration.
  • Primary and secondary choices are deliberate. List the values you expect to retain, including owner, lifecycle stage, company and communication preferences.
  • Connected systems are considered. Identify integrations or internal sheets that reference either old record ID.
  • The reviewer approves the exact pair. Record the reviewer, date and rationale before execution.

As checked on August 31, 2026, merging individual records requires Merge records or Super Admin permission. Its current documentation says a merge creates a new unique Record ID, combines activities and associations, and applies property-specific rules. It also describes workflow unenrollment behavior. Review those exceptions before acting; do not assume every field follows a simple “primary wins” rule. HubSpot merge behavior

Merge duplicates only after the pair is approved

For an individual merge, open the contact, use Actions > Merge, select the other record and review the values to retain. Confirm the primary record and property choices before submitting. Follow the current HubSpot merge instructions; do not use an export as a promise that you can undo the action.

Our recommendation is to pause at that final confirmation and compare it with the approved decision log. Does the selected contact match the intended person? Are you preserving the expected owner and relationships? If a new conflict appears, stop and investigate it. The person performing the merge should not have to guess which values the reviewer intended to keep.

Validate the merged contact record and HubSpot workflows

Open the merged record and compare it with your approved expectations. Check the selected values, associated company and deals, relevant activity history, ownership and communication preferences. Confirm that downstream references still resolve as intended.

Keep the original IDs in your review log alongside the resulting ID. When an integration owns a field, watch the next normal sync before declaring the issue closed. If Salesforce is connected, review the documented sync consequences as part of that check rather than assuming both previous records will continue syncing. A technically successful merge can still expose an unresolved source-system rule.

For broader record and automation issues, use the HubSpot troubleshooting guide as an adjacent diagnostic starting point. Keep duplicate handling scoped to identity and merge validation.

We recommend checking active HubSpot workflows separately from the property comparison. Record the automation outcome you expect and inspect the resulting history. A clean-looking contact does not establish that its next task or routing action is correct.

Bulk merging needs a narrower rule, not less review

HubSpot currently requires Data Hub Professional or Enterprise to bulk manage duplicates. Individual-pair management and bulk merging have different eligibility requirements. Verify the feature in your HubSpot account before planning the cleanup. HubSpot bulk duplicate management

We would start with a small, approved sample and review every exception before expanding the batch. Save the matching rule, retained-property decisions and result checks. Keep uncertain pairs out of the batch. A rule that works for two verified email addresses belonging to one contact may be unsuitable for a shared mailbox.

If you evaluate an app from the HubSpot Marketplace to automate this job, require a demonstration using your synthetic test cases. Compare access permissions, merge rules, exception handling and the audit record. Do not assume third-party solutions follow native behavior, or apply contact rules unchanged to companies, deals or custom objects.

Separate the merge decision from communication permission

A confirmed identity match does not settle which messages the person should receive. In your review notes, record the relevant subscription information and any uncertainty separately from the decision that two records represent the same person. Do not treat a merge as permission to resume outreach.

For example, imagine the confirmed work-and-personal-email pair in the table above. Your team may have a valid reason to consolidate the CRM history while still needing to investigate the communication preferences attached to each address. Hold any uncertain sending decision for the responsible owner. This is a proposed review practice, not a statement that HubSpot applies one universal preference rule to every merge.

Illustrative meeting room with review notes and a dashboard presentation
Editorial images are AI-generated illustrations. Screens and figures do not represent client results or verified HubSpot interfaces.

Frequently asked questions

Can I merge contacts just because their names match?

No. Matching names are a reason to investigate, not sufficient identity evidence. Confirm the person using reliable identifiers and relevant history, then review the properties and relationships that would be affected.

Does exporting the records make a merge reversible?

No. An export can preserve selected evidence for comparison, but it is not a complete rollback. HubSpot documents that records cannot be unmerged. Review its current merge rules before an authorized user proceeds.

Should I merge every candidate in a duplicate queue?

No. Separate confirmed duplicates, confirmed distinct people and unresolved candidates. Preserve the reason for each decision. Ambiguous candidates should remain open for investigation rather than being combined to make the queue smaller.

How do I know the prevention rule worked?

Check the actual record outcome from an approved repeat-entry test and inspect the next normal integration cycle. Confirm identity, values and relationships. A successful test of one import does not establish that forms or other integrations use the same matching behavior.

Prevent duplicates in imports and form submissions

Review how HubSpot contacts enter through imports, form submissions, integrations and manual entry. For each route, document the matching key, required data and exception owner. Test the same synthetic person twice through an approved test path and inspect whether your intended update behavior occurs. Do not use production contacts for an uncontrolled experiment.

Before the next import, we recommend cleaning the source file, standardizing input fields and checking the chosen unique identifier against existing records. Keep a sample that should update one contact and another that should create a new contact. This tests your intended behavior without treating every matching name as a duplicate.

Track reviewed candidates, approved merges and recurring creation paths separately. A smaller candidate queue is useful, but it does not prove your identity rules are correct.

If duplicate records are part of a wider cleanup effort, begin with the portal audit checklist. For help defining the review and prevention process, request a scoped review.