Every RevOps team knows their CRM data needs work. The question is where to start. A structured audit gives you the answer: a clear picture of what is broken, how bad it is, and what to fix first.
This guide walks through a complete CRM data audit in 6 steps. It is designed for RevOps teams running Salesforce or HubSpot, but the framework works with any CRM.
Key Takeaways
- A CRM audit covers 6 dimensions: completeness, accuracy, consistency, duplicates, freshness, and connectivity.
- 76% of CRM entries are incomplete. This is almost always the biggest finding.
- 15-25% of CRM records are duplicates. Each duplicate splits engagement history and confuses routing.
- Run monthly spot checks and quarterly deep audits. Data decays too fast for annual audits to catch.
- The audit is diagnostic, not the fix. Use the audit to prioritize, then fix with enrichment and prevention.
Step 1: Define your critical fields
Before you audit, decide what matters. Not every field in your CRM is equally important. Pick the 8-10 fields that your scoring, routing, AI, and reporting actually depend on.
For most B2B companies, the critical fields are:
- Company name and domain
- Industry / vertical
- Employee count / company size
- Annual revenue
- Country / region
- Contact email (verified)
- Contact phone (direct dial)
- Job title / seniority
- Technology stack
- Funding stage
Step 2: Measure completeness
For each critical field, run a report showing the percentage of records where the field is populated. This is your completeness score.
According to CRM data hygiene research, 76% of CRM entries are less than half complete. Your audit will likely confirm this pattern.
Benchmarks:
- 95%+ populated: Excellent
- 80-95%: Acceptable
- 60-80%: Needs attention
- Below 60%: Critical, your scoring and routing are unreliable
Document each field and its completeness percentage. This becomes your priority list for enrichment.
Step 3: Check accuracy
Completeness tells you if the field is populated. Accuracy tells you if it is correct. A field filled with "Unknown" or "N/A" is populated but not accurate.
To spot-check accuracy:
- Pull a random sample of 100 records
- Verify each critical field against an external source (company website, LinkedIn, etc.)
- Count the errors
- Extrapolate to your full database
According to data accuracy research, the average B2B data provider delivers around 50% accuracy. Your CRM data, which entered via manual entry and list imports, is likely in the same range.
Step 4: Find duplicates
Duplicates are the silent killer of CRM data quality. They inflate account counts, split engagement history, confuse lead routing, and make reporting unreliable.
To find duplicates:
- Match by email domain (all contacts at the same domain should be on the same account)
- Match by company name (watch for variations like "Acme Inc" vs "Acme" vs "ACME Inc.")
- Match by phone number
Most CRMs have 15-25% duplicate rates. That means if you have 50,000 contacts, 7,500 to 12,500 are duplicates that need to be merged or deleted.
Step 5: Measure freshness
Check when critical fields were last updated. Data that has not been refreshed in 6+ months is likely stale.
According to CRM data quality benchmarks, B2B contact data decays between 22.5% and 70.3% annually, with email decay at 3.6% monthly.
Run a report showing the percentage of records where key fields were last modified more than 90, 180, and 365 days ago. Any record untouched for 365+ days is almost certainly stale.
Step 6: Check connectivity
Connectivity measures whether records are properly linked. Contacts should be attached to accounts. Accounts should have associated deals. Deals should have activities.
Orphaned records (contacts not attached to accounts, accounts with no contacts) break reporting and AI workflows. They are invisible to most reports and create blind spots in your pipeline.
After the audit: what to do with the results
Your audit gives you a score across 6 dimensions. The fix depends on what you find:
- Completeness below 80%: Bulk enrich from an external data source like Landbase (1,500+ fields per account)
- Accuracy below 80%: Re-verify your top accounts against current data
- Duplicates above 10%: Run a deduplication pass with merge rules
- Freshness: records older than 6 months: Schedule quarterly re-enrichment for your top accounts
- Connectivity issues: Run a cleanup to link orphaned contacts to accounts by domain
The most important change is prevention: set up enrichment at the point of entry so new records arrive complete, accurate, and current. This prevents the problems from growing after you fix them.
Frequently asked questions
How long does a CRM data audit take?
A basic audit (completeness + duplicates) takes 2-4 hours. A full audit across all 6 dimensions takes 1-2 days. The time depends on your CRM size and how easy it is to run reports.
Can I automate the audit?
Partially. Completeness and duplicate reports can be automated as saved reports or dashboards. Accuracy spot-checks require manual verification. Freshness can be automated with last-modified-date reports.
What tools do I need for a CRM audit?
Your CRM reporting (Salesforce Reports or HubSpot Reports) handles most of it. For deduplication, tools like Cloudingo or DemandTools help. For enrichment to fix the gaps, Landbase delivers verified accounts with 1,500+ enrichment fields as a CSV export.
Should I audit before or after deploying AI tools?
Before. Always before. AI amplifies whatever data it receives. Deploying AI on dirty data produces bad output at scale. Run the audit, fix the critical gaps, then deploy AI on clean inputs.
Build a GTM-ready audience
.png)
.png)
