Duplicate companies in HubSpot keep coming back

Where they actually come from

Four sources cover almost every case. Form submissions matched on email address rather than company domain. Spreadsheet imports run without a deduplication rule. Enrichment or prospecting tools creating records when they cannot match. And reps typing a name slightly differently because searching felt slower than creating.

Why merging alone fails

A merge is a point in time. The rules that allowed the duplicate are still running the day after. Teams that merge without governance see the duplicate rate climb back to roughly where it was within a quarter, which is how organisations end up doing this every year and concluding that CRM hygiene is hopeless.

What blocking actually looks like

Normalise on domain rather than company name. Require a domain on creation from forms and integrations. Set matching rules on import rather than trusting the file. Give reps a search-first flow that makes finding the account easier than creating one. Then run a scheduled merge for the backlog rather than a heroic one-off.

The cost of leaving it

Duplicates do not just look untidy. They split the activity history, so nobody can see that this account already said no. They break attribution, because revenue lands on one record and the campaign on another. They inflate counts, so your pipeline and your TAM are both wrong. And they teach the sales team that the CRM is not the truth.

Count these before you merge anything

• Accounts sharing a domain but not a record
• Records created in the last 90 days with no domain at all
• Which integrations have create permission
• Whether your forms match on email or on domain