Roughly 30% of B2B contact data decays every year. For a UK sales team with 10,000 CRM records, that means 3,000 contacts quietly become unreachable, misattributed, or flat-out wrong between January and December. The cost is not theoretical. It shows up in bounced sequences, wasted rep hours, and pipeline that looks full but never closes.
The problem is that "bad data" is vague. Leaders know their CRM is not perfect, but they cannot pin down where or why. That is where the six dimensions of data quality come in. Each one is specific, measurable, and directly tied to revenue impact.
The Six Dimensions of CRM Data Quality
Data quality is not a single score. It is six distinct measures, each of which can fail independently. A CRM can have accurate email addresses but still fall apart on completeness or timeliness. Understanding each dimension is the first step to diagnosing which ones are costing your team money.
1. Accuracy
Accuracy answers one question: does the data reflect reality? If a contact's job title says "Marketing Manager" but they were promoted to VP six months ago, the record is inaccurate. If the company headcount says 50 but they have grown to 200, same problem.
Inaccurate data causes misrouted leads, irrelevant messaging, and wasted outreach. An SDR who calls a prospect using an old direct dial wastes three to five minutes per attempt. Across a team of eight reps making 50 calls a day, even a 10% inaccuracy rate burns 20 to 40 minutes of selling time daily.
Accuracy degrades fastest with job titles and phone numbers. People change roles every 2.5 years on average, and direct dials change even more frequently when companies switch phone systems or move offices.
2. Completeness
Completeness measures whether the fields that matter are actually populated. A contact record with a name and email but no phone number, no company, and no job title is technically in your CRM but practically useless for outbound.
Most B2B CRMs have completeness rates between 40% and 60% on fields that reps actually need: direct dial, LinkedIn URL, company revenue, and industry vertical. The gap is not because the data does not exist. It is because nobody enriched the records after the initial import or form submission.
Incomplete records create a two-tier database: contacts you can actually work and contacts you cannot. The second group sits there consuming storage and inflating vanity metrics while contributing nothing to pipeline.
3. Consistency
Consistency means the same data follows the same format everywhere. "United Kingdom" in one record, "UK" in another, and "Great Britain" in a third. "SaaS" versus "Software as a Service" versus "Software-SaaS". Phone numbers with country codes in some records and without in others.
Inconsistent data breaks segmentation, reporting, and automation. A workflow that triggers on industry equals "SaaS" misses every record labelled "Software as a Service". A territory report that filters by country equals "UK" undercounts every record stored as "United Kingdom".
The root cause is almost always a lack of standardised input rules: no dropdown constraints, no validation on form fields, no normalisation on import. Once inconsistency gets into the CRM, it compounds with every new record.
4. Timeliness
Timeliness measures how current the data is. A record that was accurate and complete 18 months ago may be neither today. People move companies. Businesses get acquired. Office addresses change.
The half-life of B2B contact data is roughly 12 to 18 months. After that point, more than half of a static database will contain at least one outdated field. For UK companies specifically, Companies House data shows that roughly 500,000 companies are dissolved, struck off, or undergo significant changes each year.
Timeliness is the dimension most teams neglect because it requires ongoing effort. A one-off cleanup fixes the other five dimensions at a point in time. Timeliness requires a recurring process - monthly or quarterly checks against live data sources.
5. Uniqueness
Uniqueness means every real-world entity is represented once. Duplicate records are one of the most common CRM problems, and one of the most expensive. A typical B2B CRM with 10,000+ records has a duplicate rate between 10% and 30%.
Duplicates cause reps to contact the same person twice from different sequences. They inflate pipeline forecasts. They break lead routing by splitting a contact's engagement history across two records, so neither triggers the right score or workflow.
Duplicates enter the CRM through multiple channels: form submissions with slight name variations, CSV imports without deduplication, integrations that create new records instead of matching existing ones. The longer they go unaddressed, the harder they are to merge because each duplicate accumulates its own activity history.
6. Validity
Validity checks whether data conforms to the rules of its field. An email address without an @ symbol is invalid. A UK phone number with only nine digits is invalid. A postcode that does not match any real location is invalid.
Invalid data typically enters through manual entry, poorly configured forms, or bulk imports without validation. It is the easiest dimension to fix because validity checks are binary: the data either meets the format rules or it does not.
But invalid data that goes unchecked cascades into other problems. Invalid emails cause hard bounces, which damage sender reputation. Invalid phone numbers waste dial time. Invalid postcodes break territory assignment.
What Poor Data Quality Costs in Practice
Each dimension has a direct financial impact. Here is a conservative model for a 10-person UK sales team:
- Accuracy: 2 hours per rep per week chasing outdated contacts = 1,040 hours per year = roughly £26,000 in lost selling time
- Completeness: 40% of records unusable for outbound = 4,000 dead records in a 10,000-record CRM, each representing a missed opportunity
- Consistency: 15% of automated workflows misfiring due to format mismatches = leads falling through cracks every week
- Timeliness: 30% annual decay = 3,000 records going stale each year without a refresh process
- Uniqueness: 20% duplicate rate = 2,000 duplicates causing double-touches and inflated forecasts
- Validity: 5% hard bounce rate on email sequences = damaged sender domain reputation and reduced deliverability across all outreach
Added together, a mid-market UK B2B team with no active CRM data hygiene programme is losing somewhere between £50,000 and £150,000 per year in wasted effort, missed pipeline, and operational friction. The exact number depends on team size, deal value, and how long the CRM has gone without attention.
How to Measure Your CRM Data Quality
You do not need a consultant to get a baseline score. Export your CRM to a spreadsheet and check these five things:
- Field completeness rate: For your 10 most important fields (email, phone, job title, company, industry, revenue, headcount, LinkedIn, city, last activity date), what percentage of records have all 10 populated?
- Email bounce rate: Run your contact list through a verification tool. Anything above 3% needs attention. Above 8% is actively damaging your sender reputation.
- Duplicate count: Search for contacts sharing the same email address or the same first name + last name + company combination. Most CRMs have a built-in duplicate finder, though they miss partial matches.
- Last modified date distribution: How many records have not been updated in 12+ months? Those are your timeliness risk.
- Format consistency: Pick one field - country, for example - and count how many unique values exist. If "country" has more than 10 variations for the UK alone, you have a consistency problem.
If you want a professional assessment across all six dimensions, a CRM health check gives you a scored report with specific fix recommendations in under a week.
Fixing the Problem
Data quality is not a one-time project. It is an ongoing operational discipline. The six dimensions give you a framework to prioritise: fix validity first (it is the easiest), then uniqueness (deduplicate), then accuracy and completeness (enrich), then consistency (standardise), and finally timeliness (set up recurring checks).
The teams that treat data quality as a continuous process rather than an annual cleanup are the ones whose pipeline numbers actually mean something. Everyone else is making decisions on data they cannot trust.
