The Complete CRM Audit Checklist: 50 Things to Check Before You Optimize

A CRM audit is not a cleanup exercise. It is how you find out whether your CRM can be trusted before you add more automation, reporting, AI, integrations, or sales process on top of it.

When a HubSpot portal feels messy, the problem is rarely one thing. It is usually a combination of weak data standards, unclear lifecycle rules, old workflows, duplicate properties, reporting shortcuts, and permissions that grew organically over time.

This is the checklist I would use before optimizing a CRM, migrating data, rolling out Breeze AI, or asking leadership to trust pipeline and marketing numbers.

1. Data model and hygiene

Start with the data. If the CRM cannot describe customers cleanly, every dashboard, workflow, handoff, and AI answer becomes less reliable.

1. Contacts and companies are deduplicated with a clear matching rule.
2. Company domains, names, countries, industries, and employee ranges are consistently populated.
3. Lifecycle stage is defined, enforced, and not manually overwritten without reason.
4. Required fields are actually required at the right moment, not only documented in a spreadsheet.
5. Custom properties have clear names, descriptions, owners, and usage notes.
6. Unused properties are archived or clearly marked as deprecated.
7. Contact, company, deal, ticket, and custom object associations are reliable enough for reporting.
8. Import processes prevent bad data instead of creating recurring cleanup work.
9. Email validity, bounced contacts, unsubscribes, and consent fields are monitored.
10. Data enrichment is mapped to source fields so teams know what came from forms, users, integrations, or enrichment tools.

2. Lifecycle, pipeline, and handoffs

This is where marketing and sales alignment either becomes operational or stays theoretical.

11. Lifecycle stage definitions are written in business language and understood by marketing, sales, and customer success.
12. MQL, SQL, opportunity, customer, and evangelist stages have objective entry criteria.
13. Lead status is not being used as a second, conflicting lifecycle system.
14. Deal stages match how sales actually works, not how the CRM was configured years ago.
15. Every deal stage has a clear exit requirement.
16. Ownership rules are documented for contacts, companies, deals, tickets, and accounts.
17. Handoff rules between marketing, sales, SDRs, account executives, and customer success are visible inside the CRM.
18. Stale leads, stalled deals, recycled leads, and disqualified records have a defined path.

3. Workflow and automation risk

Automation should scale a process that already works. It should not hide a process nobody owns.

19. Active workflows have owners, descriptions, and last-reviewed dates.
20. Critical workflows are grouped by business purpose: routing, nurture, lifecycle, sales ops, support, internal ops.
21. Workflow enrollment criteria are specific enough to avoid accidental enrollment.
22. Suppression lists and exclusion rules are documented for email and lifecycle automations.
23. Workflows that update lifecycle stages, owners, lead status, deal stages, or attribution fields are reviewed carefully.
24. Workflow conflicts are checked, especially when multiple automations update the same property.
25. Manual overrides are possible where the business needs exceptions.
26. There is a rollback plan for automations that create bad tasks, emails, records, or property updates.

4. Reporting and attribution trust

The dashboard is not the source of truth. The data underneath it is.

27. Leadership dashboards are tied to documented business questions.
28. Revenue, pipeline, MQL, SQL, conversion, source, and campaign reports all use agreed definitions.
29. Attribution fields are populated consistently and not overwritten by later imports.
30. Closed-won and closed-lost reasons are structured enough to be useful.
31. Forecast, pipeline, and stage-conversion reports exclude test data and obvious junk records.
32. Marketing reports can connect campaigns to contacts, companies, deals, and revenue where relevant.
33. Every critical dashboard has an owner who understands the fields and filters behind it.

5. Governance, permissions, and accountability

If everyone can change everything, the CRM will eventually reflect nobody's process.

34. Admin rights are limited to people who need them.
35. Property creation, workflow creation, list creation, and pipeline edits follow a review process.
36. Teams know who owns data quality, automation, reporting, integrations, and permissions.
37. Naming conventions exist for properties, lists, workflows, campaigns, reports, and dashboards.
38. Archived assets are cleaned regularly so old logic does not confuse current users.
39. GDPR, consent, subscription, and deletion processes are documented and tested.
40. Change management is visible: major CRM edits are logged, announced, and reviewed after launch.

6. Integrations and migration readiness

Most CRM problems get worse when another system starts syncing into them.

41. Every integration has a documented purpose, owner, data direction, and sync frequency.
42. Field mappings are documented for key tools such as forms, enrichment, support, billing, ads, data warehouse, and sales engagement.
43. Sync errors are monitored instead of discovered during reporting reviews.
44. Migration imports are tested on samples before full import.
45. There is a clear source of truth when two systems disagree.

7. AI readiness

AI makes CRM readiness more important, not less. If AI reads bad context or writes to weak processes, it creates faster mess.

46. AI use cases are classified as assist, recommend, write-with-review, or act automatically.
47. Knowledge sources used by AI are current, owned, and consistent with sales and support reality.
48. AI-generated customer-facing content requires review unless the answer space is narrow and tested.
49. AI workflows that update records, create tasks, send messages, or change lifecycle/reporting fields have audit trails.
50. Teams know when to escalate AI output to a human instead of trusting a polished answer.

What to do with the score

If you find a handful of issues, fix them as maintenance. If you find 10 to 25, treat it as a focused CRM cleanup project. If you find more than 25, the CRM is probably not ready for serious optimization, AI rollout, migration, or executive reporting without a structured remediation plan.

That is where a real audit earns its money. It turns "the CRM feels messy" into a prioritized map: what breaks reporting, what blocks automation, what creates sales friction, what risks compliance, and what AI would amplify.

Related reading

• Why Data Quality Is the Foundation of Every CRM Strategy
• Process Design for CRM Operations
• HubSpot AI Agents Need CRM Governance Before They Need More Prompts