How to Calculate Your Data Quality Score

A data quality score is a composite metric (0-100) that quantifies the health of your CRM data across multiple dimensions: completeness, accuracy, freshness, duplication, and consistency. Each dimension is scored independently and then combined with weights that reflect its relative impact on business operations. The result is a single number that answers the question "how good is our data?" and a breakdown that shows where to focus improvement efforts.

Most companies have a vague sense that their data is "not great." That is not actionable. A data quality score turns the vague sense into specific, measurable problems. Your completeness is 72% because 28% of contacts are missing email addresses. Your accuracy is 61% because nearly 4 in 10 job titles are outdated. Your duplication rate is 18%. These are problems you can assign, budget for, and track progress on.

This guide walks through how to measure each dimension, weight them appropriately, and calculate a composite score you can track over time.

The Five Dimensions of Data Quality

There are many ways to slice data quality, but five dimensions cover the territory that matters for B2B operations:

Dimension	Weight	What It Measures
Accuracy	30%	Is the data correct? Do email addresses work? Are job titles current?
Completeness	25%	Are required fields filled in? Email, phone, title, company, industry?
Freshness	20%	How recently was the data verified or updated?
Duplication	15%	What percentage of records are duplicates?
Consistency	10%	Are fields formatted uniformly? Phone numbers, state names, job titles?

The weights reflect real-world impact. Accurate data matters more than consistent formatting. Complete records matter more than low duplication. You can adjust these weights based on your business priorities, but this distribution works well as a starting point for most B2B companies.

How to Measure Each Dimension

1. Accuracy (Weight: 30%)

Accuracy is the hardest dimension to measure because it requires verification against external sources. You cannot tell whether a job title is correct by looking at it in your CRM. You need to check.

How to measure:

1. Pull a random sample of 200-500 records from your database
2. For each record, verify key fields against external sources (LinkedIn for job title, email validation tool for email address, company website for company info)
3. Score each field as correct or incorrect
4. Calculate the percentage of fields that are correct across the sample

Key fields to verify: email address (use a validation tool), job title, company name, phone number.

2. Completeness (Weight: 25%)

Completeness measures the percentage of records that have all required fields populated. Define your required fields first, then measure.

Recommended required fields for B2B contacts:

• Email address
• First name and last name
• Job title
• Company name
• Phone number (direct or mobile)

How to measure:

1. For each required field, count the number of records where the field is not blank
2. Divide by total records to get the fill rate per field
3. Average the fill rates across all required fields

Example: If you have 50,000 contacts and the fill rates are Email: 82%, Title: 68%, Phone: 45%, First Name: 98%, Company: 95%, your completeness score is (82 + 68 + 45 + 98 + 95) / 5 = 77.6%.

3. Freshness (Weight: 20%)

Freshness measures how recently your data was updated or verified. B2B data decays at 30% per year, so data that has not been touched in 12+ months is significantly degraded.

How to measure:

1. Pull the "Last Modified Date" for each record
2. Categorize records by age: 0-3 months (fresh), 3-6 months (aging), 6-12 months (stale), 12+ months (decayed)
3. Assign scores: Fresh = 100, Aging = 75, Stale = 40, Decayed = 10
4. Calculate the weighted average

Example: If 30% of records are fresh, 25% aging, 20% stale, and 25% decayed, your freshness score is (0.30 x 100) + (0.25 x 75) + (0.20 x 40) + (0.25 x 10) = 30 + 18.75 + 8 + 2.5 = 59.25.

4. Duplication (Weight: 15%)

Duplication is the inverse of your duplicate rate. Lower duplication means a higher score.

How to measure:

1. Count the total number of contact records
2. Count the number of unique contacts (deduplicate by email, then by name + company for records without email)
3. Calculate: Duplication Score = (Unique Records / Total Records) x 100

Example: 50,000 total records, 43,000 unique = (43,000 / 50,000) x 100 = 86. You have a 14% duplication rate and an 86% duplication dimension score.

5. Consistency (Weight: 10%)

Consistency measures whether fields are formatted uniformly. Inconsistent data breaks automations, reporting, and segmentation.

Fields to check for consistency:

• Phone numbers: Are they all formatted the same way? (123) 456-7890 vs 1234567890 vs +1-123-456-7890
• State/Country: "California" vs "CA" vs "ca" vs "Calif."
• Job titles: "VP of Sales" vs "Vice President, Sales" vs "VP Sales"
• Company names: "Acme Inc." vs "ACME" vs "Acme, Inc"

How to measure:

1. For each field, identify the dominant format (the one used most frequently)
2. Count the percentage of records that match the dominant format
3. Average across all checked fields

Calculating Your Composite Score

Once you have a score (0-100) for each dimension, calculate the composite:

Worked example:

• Accuracy: 72
• Completeness: 78
• Freshness: 59
• Duplication: 86
• Consistency: 65

Score = (72 x 0.30) + (78 x 0.25) + (59 x 0.20) + (86 x 0.15) + (65 x 0.10) = 21.6 + 19.5 + 11.8 + 12.9 + 6.5 = 72.3

Interpreting Your Score

Score Range	Rating	What It Means
90-100	Excellent	Active data governance in place. Rare for companies without a dedicated data team.
80-89	Strong	Data supports operations well. Minor issues that do not significantly impact revenue.
60-79	Needs Work	Data is usable but causing friction. Reps work around data problems daily. Campaigns underperform.
40-59	Poor	Data is actively hurting operations. High bounce rates, wasted rep time, unreliable reporting.
Below 40	Critical	Data is a liability. Major cleanup or rebuild needed before the database is operationally useful.

Tracking Your Score Over Time

A single score is useful. A trend line is powerful. Measure monthly and track how your score changes as you invest in data quality.

What to track:

• Composite score (the headline number)
• Individual dimension scores (to spot which areas are improving or declining)
• Record count (to understand if score changes are driven by data volume or quality)
• Source-level scores (if you track data sources, score each one separately to identify which sources produce the best and worst data)

When to Get Help

If your data quality score is below 60, the cleanup work required is significant. Manual measurement, deduplication, validation, and enrichment across tens of thousands of records takes weeks of focused effort. If you do not have a dedicated data team, outsourcing the cleanup and initial scoring is often faster and more cost-effective.

We help companies measure, clean, and maintain their data quality. Get in touch if you want help getting your score above 80.

Common Questions

Related: Data Quality Metrics | Data Quality Dashboards | CRM Data Quality Checklist | Data Cleaning Services

Further reading: Data Quality Roadmap | Data Quality for Sales Leaders