Most B2B teams know their CRM data is not perfect. Few know exactly how imperfect it is, or which problems are actually costing them money. A CRM data audit answers both questions - systematically, with numbers rather than gut feeling.
This guide covers the full audit process in five steps:
- Define what to audit
- Build a scoring system
- Choose your tools
- Create a report template
- Set audit frequency
By the end, you will have a repeatable framework that takes 4-6 hours per quarter and gives you a clear, comparable view of your data health over time.
Step 1: Define What to Audit
Not every field in your CRM deserves audit attention. Focus on the fields that drive revenue operations - the ones used for segmentation, routing, reporting, and outreach.
Contact-level fields:
- Email address: Is it present? Is it valid? Is it a personal or role-based address?
- Job title: Is it populated? Does it match your CRM's controlled vocabulary (if you have one)?
- Phone number: Is it present? Is it in a valid format?
- Company association: Is the contact linked to a company record?
- Owner: Is a sales rep assigned?
- Lifecycle stage: Is it set? Does it reflect reality?
Company-level fields:
- Domain: Is the company website present?
- Industry: Is it populated with a standardised value?
- Employee count: Is a size band assigned?
- Country: Is it present and standardised?
- Annual revenue: Is a revenue band assigned (even if estimated)?
Deal-level fields:
- Close date: Is it in the future for open deals? Is it realistic?
- Amount: Is a value assigned?
- Stage: Does it reflect actual progress? When was it last updated?
- Associated contacts: Is at least one contact linked?
Cross-cutting metrics:
- Duplicate rate: What percentage of records have probable duplicates?
- Stale records: What percentage of contacts have had no activity in 6+ months?
- Orphan records: How many contacts are not associated with any company?
- Bounce rate: What was the bounce rate on your last email campaign?
Step 2: Build a Scoring System
Raw numbers are useful, but a scoring system makes it easier to track progress over time and communicate results to stakeholders who do not live in the data.
Field completeness score: For each audited field, calculate the percentage of records where the field is populated with a valid value. "Valid" is important - a phone number field containing "123" is populated but not useful. Weight each field by importance. Email completeness matters more than phone number completeness for most B2B outreach models.
A simple weighting example:
- Email: weight 5
- Job title: weight 4
- Company association: weight 5
- Phone number: weight 2
- Industry: weight 3
- Employee count: weight 2
Overall health score: Calculate a weighted average across all fields to produce a single number between 0 and 100. This is your headline metric - the number you track quarter over quarter.
Grading scale:
- 90-100: Excellent. Your data is in strong shape. Focus on maintaining it
- 70-89: Good. There are gaps, but they are manageable. Prioritise the lowest-scoring fields
- 50-69: Fair. Data quality is likely affecting revenue operations. A focused cleanup is warranted
- Below 50: Poor. Your CRM is a liability rather than an asset. Comprehensive remediation is needed
Most B2B CRMs we audit at ClientWise score between 45 and 65 on their first assessment. That is not a criticism - it is simply what happens when data accumulates over years without systematic quality management.
Step 3: Choose Your Tools
You do not need expensive software to run a meaningful audit. The right tool depends on your CRM platform and the depth of analysis you need.
Built-in CRM reporting: Both HubSpot and Salesforce offer reports that can measure field completeness. In HubSpot, create a custom report filtering contacts by "[field] is known" versus "[field] is unknown." In Salesforce, use report types with cross filters or field completion metrics in the Data Quality Analysis dashboard.
Spreadsheet analysis: Export your CRM data and use pivot tables, COUNTBLANK, and conditional formatting to identify patterns. This approach is slower but gives you maximum flexibility for custom analysis. It works well for CRMs under 50,000 records.
Dedicated tools: Insycle, Validity DemandTools, and RingLead offer data quality dashboards with pre-built audit metrics. These are worth the investment if you plan to run audits regularly and want automated duplicate detection and trend tracking.
Custom scripts: For teams with technical resources, a Python script pulling from your CRM's API can generate highly customised audit reports. This is the approach we use at ClientWise - it allows us to build client-specific quality rules and automate the entire audit pipeline.
Step 4: Create a Report Template
Consistency matters more than sophistication. A simple template that you use every quarter is infinitely more valuable than a comprehensive one you use once.
Report sections:
- Executive summary: Overall health score, change since last audit, top three issues, recommended actions
- Field completeness breakdown: Each audited field with current percentage, previous percentage, and trend direction
- Duplicate analysis: Estimated duplicate count, duplicate creation rate since last audit, and top duplicate clusters
- Stale record analysis: Count and percentage of records with no activity in 6+ months, segmented by lifecycle stage
- Recommendations: Prioritised list of actions, each with estimated effort and expected impact
Visualisation: A simple traffic light system (red/amber/green) for each metric makes the report immediately scannable. Stakeholders who receive a wall of percentages will not read it. Stakeholders who see three red indicators will ask questions.
Keep the report to two pages maximum. If you need more detail for specific remediation work, attach it as an appendix.
Want a professional audit without the setup work? Our CRM Quality Audit delivers a comprehensive data quality report within 48 hours - including duplicate analysis, completeness scoring, and prioritised recommendations.
Step 5: Set Audit Frequency
A single audit is a snapshot. Regular audits create a trend line - and trends are where the real insight lives.
Quarterly audits: This is the right cadence for most B2B teams. It is frequent enough to catch degradation before it becomes severe, and infrequent enough to be sustainable. A well-structured quarterly audit takes 4-6 hours once the template and tools are in place.
Monthly spot checks: Between full audits, run a quick check on your three most important metrics. This takes 30 minutes and serves as an early warning system. If email bounce rate suddenly jumps from 1.5% to 4%, you want to know in weeks, not months.
Trigger-based audits: Certain events should prompt an immediate data quality review:
- After a bulk import of 500+ records
- After a CRM migration or major integration change
- After a team restructuring that reassigns record ownership
- After any campaign with a bounce rate above 3%
Annual deep dive: Once a year, expand the audit to include fields beyond your usual scope. Review custom field usage (which fields are actually populated and used?), integration data quality (which connected tools are introducing bad data?), and process compliance (are teams following the data entry standards you have set?).
What Your First Audit Will Probably Reveal
Having audited hundreds of B2B CRMs, we see the same patterns repeatedly:
- Email validity is worse than expected. Teams assume 2-3% invalid emails. The real number is usually 8-15%, especially for databases older than two years
- Duplicate rates cluster around 10-15%. Most are not exact duplicates - they are near-matches caused by inconsistent data entry
- Job title completeness is low. It is typically the worst-performing field, sitting at 40-60% completeness for most CRMs
- Company association gaps are common. 5-10% of contacts are orphans with no linked company record
- Deal data is surprisingly poor. Close dates in the past for "open" deals, amounts of zero, and stages that have not been updated in months
None of this is unusual. The point of auditing is not to achieve perfection - it is to know where you stand and improve systematically. A team that moves from a health score of 52 to 71 over four quarters has materially improved its revenue operations, even though 71 is far from perfect.
For a broader framework on building sustained data quality practices, see our RevOps guide to improving CRM data quality. If your audit reveals problems you already suspected, our analysis of how CRM data kills pipeline quantifies the revenue impact.
When was the last time someone measured the actual quality of your CRM data?
