The HubSpot data audit checklist: 40 checks for contacts, companies and deals

This checklist lists the 40 checks Forgeo runs on a HubSpot portal: 14 on contacts, 14 on companies and 12 on deals. You can run each of them by hand with a filtered view in HubSpot, or let Forgeo run all of them after every sync.

How to use this checklist

Each check has a severity:

• High: the record is unreliable or blocks your team, like a duplicate or an open deal with no close date.
• Medium: the record works, but a key field is missing, like the owner.
• Low: useful information is missing, like a job title or a city. Worth filling, not urgent.

To run a check in HubSpot, open the object's index page, add a filter on the property given in brackets, and choose is unknown for an empty field. The record count tells you how big the problem is.

Forgeo turns the high and medium checks into a data quality score for each object: the share of records with no high or medium issue. A contact with only a missing job title still counts as clean. The overall score is the average of the objects' scores, weighted by their number of records.

Contacts: 14 checks

High:

• Duplicate contacts. Same email, or same name with at least two other signals in common: company, phone or LinkedIn profile.

Medium:

• Missing first name (firstname): empty, or with no letters at all, such as a phone number typed into the first name.
• Missing last name (lastname): empty, or with no letters.
• No owner (hubspot_owner_id).
• No activity for more than 12 months: last activity (notes_last_updated), or the create date when there is none, older than twelve months.

Low:

• Missing phone number: neither phone nor mobilephone.
• Missing job title (jobtitle).
• Missing lifecycle stage (lifecyclestage).
• Missing lead status (hs_lead_status).
• Missing LinkedIn profile (hs_linkedin_url).
• Missing company name (company).
• Missing city (city).
• Missing country (country).
• Possible duplicates. Same name and the same company, phone or LinkedIn profile, or the same LinkedIn profile alone. Review these before merging.

Companies: 14 checks

High:

• Duplicate companies. Same domain (protocol, "www." and path ignored), or same name with at least two other signals in common: LinkedIn page, phone or city.

Medium:

• No website: neither website nor domain.
• No associated contact: nobody is attached to the company.
• No owner (hubspot_owner_id).

Low:

• Missing domain (domain): a website is set but not the domain, which HubSpot uses to deduplicate companies.
• Missing LinkedIn page (linkedin_company_page).
• Missing lifecycle stage (lifecyclestage).
• Missing size (numberofemployees).
• Missing industry (industry).
• Missing description (description).
• Missing city (city).
• Missing founded year (founded_year).
• Missing phone (phone).
• Possible duplicates. The same LinkedIn page alone, or the same name with the same phone or city.

Deals: 12 checks

High:

• Open deal with no next activity (notes_next_activity_date).
• Open deal without a close date (closedate).
• No associated contact.
• Missing amount (amount).
• Won deal without an amount.
• Open deal with no activity for 30 to 89 days.
• Open deal with no activity for 90 days or more.

Medium:

• No next activity, whatever the stage.
• No associated company.
• Open deal with no activity for 15 to 29 days.

Low:

• Lost deal with no next activity: no follow-up planned.
• Missing original source (hs_analytics_source).

The three inactivity checks don't overlap: a deal is counted in one band only, based on its last activity, or its create date when there is none.

What to fix first

1. Duplicates, so you don't fix the same record twice.
2. High deal checks: close dates, amounts, next activities and contacts drive your forecast.
3. Owners, on contacts and companies: a record without an owner has no one to keep it up to date.
4. Empty fields you can fill from your own data, such as a domain from the website.
5. Low checks, through enrichment, as they matter for segmentation and outreach.

Forgeo runs these 40 checks after each sync and links every check to the records it found, in a spreadsheet view you can search, sort and filter.