Most teams discover their CRM is broken the day a forecast misses. By then there are usually 6 months of bad data underneath the dashboard. The fix is a 60-minute walkthrough you can run quarterly. Open HubSpot, set a timer, and follow the five steps below.
Step 1 — Pull a snapshot of contacts and flag duplicates
Export every contact created in the last 6 months. Sort by email domain, then by phone. In a fast-growing pipeline expect 18–23% of rows to share an email or phone with another contact. Each duplicate breaks lifecycle stage, breaks segmentation, and quietly skews every email-engagement metric you use to plan campaigns. Dedupe first, then audit — otherwise the rest of the numbers are wrong.
Step 2 — Find contacts not linked to a deal
Filter the contact list for "Number of associated deals = 0" where Lifecycle Stage is Customer or Opportunity. These are leaks: people who paid you but were never tied to revenue. We routinely see brands with 340+ orphaned contacts representing roughly €11,200 in untracked deal value. Re-associate them in batches of 50 with the bulk-edit tool.
Step 3 — Surface stale sequences and unsent emails
Open Marketing → Email and filter by status = "Failed to send" plus "Stuck in queue". One scan typically finds 23 unsent emails — campaign reminders, abandoned-cart triggers, win-back flows. Each one is a recoverable touchpoint worth €18–€40 in expected revenue. Look at the failure reason in the side panel: 70% of the time it's a missing token ({{firstName}}) on a contact created via the Shopify integration.
Step 4 — Audit field hygiene on the top 100 deals
Sort deals by amount, descending, and pull the top 100. Check four required fields: deal owner, close date, source, and amount. Missing values mean wrong forecasts. Average miss rate: 31% of high-value deals have at least one blank required field. Fix the top 30 by hand; queue the rest for an enrichment workflow.
Step 5 — Spot the workflow you stopped trusting
Workflows → Active. Look for any workflow with 0 enrollments in the last 14 days. Either it's broken or unneeded. Both cost you mental overhead. Open each one and check the trigger: a "contact created with property X" trigger silently fails when the source integration stops sending property X.
What an hour of audit usually surfaces
A typical hour returns:
- 340 duplicate contacts
- 14 deals with missing owners
- 23 unsent emails worth retrying
- 8 stuck workflows
- €5,200–€11,000 of recoverable deal value
The audit takes 60 minutes. It also goes stale within 6 weeks — usually faster, because the integrations keep changing. Manual audits find problems once; automated monitoring finds them continuously, before the next forecast meeting.
