Why fake-company records pollute everything they touch
HubSpot's auto-association is a mostly-good feature: a contact with jane@acme.com gets auto-attached to a company with domain acme.com. If no Acme company exists, HubSpot creates one. Useful for real domains.
The failure mode is when the email's domain isn't a company. A contact who registers with jane@gmail.com triggers HubSpot to create — or attach to — a company called "Gmail" with domain gmail.com. There is now a fake "Gmail" record in your CRM, possibly with hundreds of unrelated contacts attached to it.
The damage:
- Account-level reporting becomes meaningless. "Gmail" appears in your top-accounts list because it has the most contacts. None of those contacts are Gmail employees.
- ABM workflows misfire. A workflow that targets the largest accounts treats Gmail as a strategic account.
- Enrichment tools get confused. Clearbit looking up
gmail.comreturns Google's data, not your actual customer's. You enrich the wrong record. - Filtering becomes painful. Every "by company" segmentation has to remember to exclude these fake records.
What two pattern types the filter actually catches
Two filter conditions in OR:
- Domain matches a personal-email pattern:
gmail.com,yahoo.com,hotmail.com,outlook.com,icloud.com,aol.com,protonmail.com, etc. - Domain matches a test pattern:
test.com,example.com,localhost,127.0.0.1,mailinator.com, etc.
Both should be archived (not deleted — HubSpot's archive preserves the records). The contacts that were attached to them need to be detached and either re-attached to real companies or left as company-less.
Why the cleanup is recurring work, not a one-time pass
The personal-domain list is open-ended:
ISP-specific email domains. comcast.net, verizon.net, bt.com — internet service provider domains used for personal email. These create fake-company records of small ISPs that look like B2B accounts.
Country-specific personal domains. mail.ru, yandex.com, gmx.de, web.de, 163.com, qq.com. Localized personal-email providers. Easy to miss if you build the blocklist from a US-centric list.
University email domains. mit.edu, harvard.edu, etc. These are real organizations but typically not commercial buyers — they're students or researchers. Whether to treat them as fake-company records depends on your motion.
The deeper friction is the cleanup itself. Detaching contacts from a fake "Gmail" record and re-attaching to their actual employers requires knowing where each contact actually works — which often requires email enrichment or LinkedIn lookup. For 10 contacts this is a coffee break. For a portal with 5,000 contacts attached to fake companies, it's a quarter-long project that nobody ever finishes.
The pragmatic approach: archive the fake-company records, leave the contacts company-less, and use enrichment going forward to gradually re-attach. Most contacts will eventually get the right company through new form fills or LinkedIn-import data.
The manual HubSpot recipe
Two filter groups (personal-domains + test-domains), sorted by associated-contact count. Maintain the domain list quarterly as new patterns emerge.
- Open Companies → Create viewNavigate to Companies → Companies. Click 'Create view' in the top right.
- Add filter: Domain is in personal-email listFilter by Company properties →
Domain→ 'is any of' →gmail.com,yahoo.com,hotmail.com,outlook.com,icloud.com,aol.com,protonmail.com,live.com,me.com,mac.com. - Add OR filter: Domain is in test/disposable listOR group →
Domain→ 'is any of' →test.com,example.com,localhost,mailinator.com,yopmail.com,tempmail.com. - Add column: Number of associated contactsCritical for triage — a fake 'Gmail' company with 800 attached contacts is the biggest cleanup; one with 2 contacts is barely worth touching.
- Sort by Number of associated contacts, descendingBiggest fake companies at the top. Triage starts there.
- Save as 'Companies — personal/test domain'Pin to your sales-ops dashboard. Archive (don't delete) the fake records — preserves history while removing them from active reporting.
What Bloated does instead
Fake-company records — detached, archived, and re-enriched in one workflow.
Bloated detects fake-company records across global personal-email patterns (US, EU, APAC variants) and test/disposable lists AND suggests re-attachment targets based on each contact's other email domains, LinkedIn data, or recent form fills. The cleanup pass actually reduces the count instead of just generating manual work.
domain · HubSpot company property