"Data quality" is one of those phrases that everyone in RevOps agrees is important and almost no one has operationalized. It gets raised in quarterly reviews when campaigns underperform, in pipeline reviews when attribution looks wrong, and in board meetings when the CRM numbers do not reconcile with sales' intuition. Then the crisis passes, the hygiene project gets put back on the roadmap, and the same conversation happens again next quarter.

The reason data quality stays perpetually unresolved is not a lack of will. It is a lack of operational definition. "Improve data quality" is not a task anyone can execute. But "ensure that 90% of contacts in our active campaign segments have a valid email, populated job title, and company size on file before any send" — that is a task with a clear acceptance criterion, an owner, and a mechanism for enforcement.

This guide gives you the framework to make data quality operational in HubSpot: the seven dimensions, how to score your current state, how to set thresholds that mean something, how to build quality gates into workflows, and how to make the whole thing a team habit rather than a quarterly fire drill.

The Seven Dimensions of CRM Data Quality

Most discussions of data quality conflate several distinct problems into a single bucket. Separating them is important because each dimension has a different cause, a different remediation approach, and a different set of metrics.

1. Completeness

Completeness measures whether the fields that should be populated are actually populated. A contact record with a name and email but no job title, company, or industry is incomplete. Completeness is the most visible dimension and the easiest to measure — it is simply the percentage of required fields with a non-null value.

HubSpot example: Create a custom report showing the percentage of contacts in your target segment with each critical field populated. This is your completeness score. Most teams find they are operating at 50-70% completeness on their critical fields — meaning 30-50% of their database is missing at least one field required for effective routing, scoring, or personalization.

2. Accuracy

Accuracy measures whether the values that are populated are actually correct. A contact with a job title of "Director of Marketing" who is now a VP of Sales has 100% completeness on the job title field and 0% accuracy. Accuracy is harder to measure than completeness because it requires either external verification or freshness inference — you cannot determine from inside the CRM whether a field value is correct without comparing it to ground truth.

HubSpot example: Accuracy is best assessed through enrichment — running a contact through a data provider and comparing the returned value to the stored value. A high enrichment disagreement rate on a field suggests low accuracy. Alternatively, bounce rates and sequence reply rates serve as lagging indicators of email accuracy specifically.

3. Freshness

Freshness measures how recently a record's data was verified or updated. A fully complete and accurate record from 18 months ago has a freshness problem — the accuracy may have decayed since the data was last confirmed. Freshness is a function of time and field-specific decay rates.

HubSpot example: Track the "Last Enriched Date" as a custom property on every contact record. Create segments for contacts not enriched in the last 90, 180, and 365 days. Contacts in the 365+ day bucket should be treated as high decay risk and prioritized for re-enrichment before any outbound motion.

4. Validity

Validity measures whether field values conform to the expected format and data type — regardless of whether they are accurate. An email address of "john@" is invalid. A phone number of "555" is invalid. A country field populated with "US" when the expected format is "United States" is technically invalid, even if everyone knows what it means. Validity problems are often introduced at the point of data entry or through inconsistent import formats.

HubSpot example: HubSpot's property validation rules catch some format issues at form submission (email syntax, URL format), but they are limited for custom properties. Regular data audits using exports and basic pattern matching will surface validity issues that slip through.

5. Consistency

Consistency measures whether the same data is represented the same way across records and across related objects. A contact record that shows "Technology" as industry and whose associated company record shows "Software" as industry has a consistency problem. A deal record that shows a different company association than the contact's primary company has a consistency problem.

HubSpot example: Cross-object consistency issues are common after integrations with external systems (Salesforce syncs, data imports) that use different taxonomy for the same field. Run consistency checks on industry, company size, and territory fields across contact and company objects periodically.

6. Uniqueness

Uniqueness measures the absence of duplicate records. A database with 10,000 contacts that actually represents 8,500 unique people has a uniqueness problem — and the 15% duplicate rate means every segment, attribution model, and sequence is working against a corrupted dataset.

HubSpot example: Uniqueness is the dimension most directly addressed by deduplication tooling. It is also one of the most measurable — you can quantify your estimated duplicate rate through a dedup tool scan and track it over time as a KPI.

7. Enrichment Coverage

Enrichment coverage measures the percentage of records that have been supplemented with third-party data — firmographics, technographics, demographic signals. A contact record populated only with what the contact self-reported at form fill has different enrichment coverage than one that has been through a multi-provider waterfall and has 20 populated fields.

HubSpot example: Track enrichment coverage as a composite score: the percentage of a defined set of enrichment fields (industry, employee count, revenue, technology stack, direct phone, seniority level) that are populated on each record. A record with 4 of 8 enrichment fields populated has 50% enrichment coverage.

How to Score Your Current Database: A Manual Audit Approach

With the seven dimensions defined, you can build a baseline score. Here is a practical approach that does not require specialized tooling.

Step 1: Define your critical field set

Start by listing the fields that are required for your primary use cases. For most RevOps teams, this includes:

Email address (required for all outbound motion)
Job title (required for scoring and personalization)
Company name (required for firmographic routing)
Industry (required for segmentation)
Company employee count (required for tier-based routing)
Country/region (required for territory assignment)
Phone number (required for outbound sequences with calling)

This is your "critical field set." You may have 7-12 fields, depending on your tech stack and motion.

Step 2: Run completeness by dimension

For each field in your critical set, query HubSpot for the percentage of your active contact population (lifecycle stage = Marketing Qualified Lead, SQL, Customer, or whatever your active statuses are) that has a non-null value. Export these numbers. This is your completeness score.

Step 3: Estimate freshness exposure

Filter your active contact population by "Last Enriched Date" (if you track it) or by "Last Activity Date" as a proxy. What percentage has not had any data update in over 12 months? Apply your known decay rates to estimate accuracy exposure. If 40% of your active contacts have not been enriched in 18 months, and job title decays at 65.8% per year, you can estimate that roughly 50%+ of that cohort has an inaccurate job title.

Step 4: Spot-check accuracy

Pull a random 50-contact sample from your active database. Manually verify each record against LinkedIn and the company website. How many job titles are accurate? How many companies still exist as described? How many emails are deliverable (you can test a sample with Hunter.io's email verification)? This accuracy spot-check gives you a confidence interval for the freshness estimates from Step 3.

Step 5: Calculate a composite score

Weight your dimensions based on their importance to your specific revenue motions. A rough starting framework:

Completeness (critical fields): 30%
Freshness: 25%
Accuracy (estimated from spot-check): 20%
Validity: 10%
Uniqueness: 10%
Enrichment coverage: 5%

Score each dimension on a 0-100 scale based on your measurements. Multiply by the weights. This is your composite HubSpot data quality score.

Most teams who run this exercise for the first time discover their composite score is between 45 and 65. A score below 60 indicates that data quality is a material constraint on revenue performance. A score above 80 indicates a high-functioning data operation.

Setting Quality Thresholds That Mean Something

Thresholds only matter if they are tied to business actions. "We want 80% field completeness" is a goal, not a threshold. A threshold specifies the minimum standard required before a business action can occur.

Campaign-ready standard: Before a contact is eligible for a campaign send, what fields must be populated? A reasonable standard for most B2B email campaigns:

Email address: present and valid (not hard bounced)
Company name: present
At least one of: job title, industry, or seniority level

This threshold prevents the worst-case scenario — personalized campaigns sent to records with no reliable fields to personalize on.

Sequence-ready standard: Before a contact is eligible for a sales sequence, what fields must be populated? A reasonable standard for outbound sequences with personalization:

Email address: present and valid
Job title: present and under 12 months since last verified
Company name: present
Industry: present (for industry-specific messaging variants)

Routing-ready standard: Before a contact is passed to a specific sales rep or territory, what firmographic fields must be populated?

Company employee count: present (for tier-based routing)
Country/region: present (for territory routing)
Industry: present (for segment-based routing)

Scoring-eligible standard: Before a contact's lead score is used to trigger pipeline actions, what completeness and freshness criteria must be met? This is the most important threshold for organizations with automated MQL triggers, because a contact that scores above the MQL threshold based on stale data can create a false MQL that wastes sales capacity.

The key principle: thresholds should prevent false positives (bad records passing quality gates and creating work for downstream teams) rather than simply setting aspirational targets.

Building Quality Gates Into HubSpot Workflows

Once you have defined your thresholds, you can encode them as workflow conditions that act as quality gates.

The Basic Gate Pattern

Create a HubSpot Workflow triggered by the lifecycle stage change that precedes the action you want to gate (e.g., triggered when "Lifecycle Stage is set to Marketing Qualified Lead"):

Check enrollment condition: Is [critical field] known? Is [email] not bounced? Is [enriched date] less than 365 days ago?
If YES: Proceed with the downstream action (campaign enrollment, sales routing, sequence enrollment)
If NO: Branch to a remediation path — trigger enrichment, set a "Data Quality Hold" flag, add to a review list for manual cleanup, or suppress from the downstream action

Enrichment-Triggered Remediation

Build a workflow that fires automatically when a record fails a quality gate:

Record fails campaign-ready check (missing job title or bounced email)
Workflow triggers enrichment via webhook to your enrichment provider
Re-check quality gate after enrichment attempt
If enrichment resolves the gap, enroll in original campaign
If enrichment does not resolve, flag for manual review

This pattern converts data quality checks from passive to active — instead of simply suppressing bad records, the workflow actively tries to fix them before suppression.

Quality Score Property

Create a custom HubSpot property called "Data Quality Score" (numeric, 0-100) and populate it through a calculated property or periodic workflow. Use this score as an enrollment criterion for high-value campaigns. Contacts must have a quality score above your threshold to be eligible.

Decay-Based Re-Enrichment

Create a workflow triggered when "Last Enriched Date is more than 365 days ago" for active contacts. This fires automatic re-enrichment, ensuring that contacts actively used in your revenue motion do not age into inaccuracy without intervention.

Making Data Quality a Team Habit (Not a Quarterly Fire Drill)

The structural problem with data quality is that it benefits the whole team but the cost of maintaining it falls on a few people (the HubSpot Admin, the RevOps analyst). If data quality is a RevOps-only concern, it will always be under-resourced relative to demand.

Making data quality a team habit requires making its costs and benefits visible to the people who create data as well as the people who clean it up.

Make data quality visible in shared dashboards. Add a "Database Health" widget to the RevOps dashboard that shows current field completeness percentages on critical fields, duplicate rate trend, and contacts currently on quality hold. When sales and marketing can see the data quality numbers as part of their regular operating cadence, they develop a stake in the outcome.

Attribute revenue to data quality. When a deal closes and the contact record was enriched, flag that in the deal record. When a campaign outperforms, check whether the segment had above-average data quality scores. Building a correlation between data quality and revenue outcomes makes the business case concrete and repeatable.

Give sales reps ownership of record accuracy. Build a HubSpot Sales task or sequence step that fires after the first call: "Update contact job title, confirm direct phone number, verify company size." Make accurate record updating a standard part of the post-call workflow, not an optional administrative task.

Establish a formal data steward role. In organizations larger than 50 seats, a formal data steward — someone who owns data quality as a defined responsibility, not a side project — makes a measurable difference. This does not need to be a full-time role; 20% of a senior HubSpot Admin's time, with a defined mandate and metrics, is sufficient for most databases under 100,000 contacts.

Review data quality monthly, not quarterly. Monthly reviews catch accumulation early. Quarterly reviews often catch problems after they have already affected a major campaign cycle. Put database health on the monthly RevOps review agenda alongside pipeline coverage and campaign performance.

From Reactive to Proactive Data Quality

The difference between reactive and proactive data quality operations is simple: reactive teams clean data after campaigns fail; proactive teams enforce quality gates that prevent bad data from reaching campaigns in the first place.

The framework in this post gives you the building blocks for a proactive data quality system:

Seven dimensions to measure, each with HubSpot-specific examples
A manual audit approach that produces a composite score in a day of work
Quality thresholds tied to specific business actions, not aspirational targets
Workflow patterns that enforce those thresholds automatically
Team habits that distribute the cost of data maintenance

The fully proactive state — where quality checks fire automatically, remediation workflows run without human intervention, and sales and marketing always work from campaign-ready records — requires tooling beyond what HubSpot's native workflows can easily support. But the structural foundation: defining dimensions, setting thresholds, and building gates into your workflow logic — is achievable today with what you already have.

What We Are Building

At MarketingSoda, we are building MarketingSoda Refine™ — a HubSpot-native data quality platform that automates this framework. The seven dimensions described in this post are encoded as our per-record quality scoring engine — every contact and company record gets an A-F grade across all seven dimensions, calculated continuously and exposed as HubSpot properties you can use in workflows, segments, and reports.

The quality gates, enrichment triggers, and remediation workflows described above are the operational layer we are building on top of that scoring: when a record's quality score drops below a threshold, enrichment fires automatically. When a campaign segment has contacts below grade, they are held until remediation completes.

We are pre-launch and building our waitlist. If this framework maps to the data quality problems you are trying to solve, we would like to have you involved in early access. Join the waitlist for MarketingSoda Refine

The RevOps Data Quality Framework for HubSpot: A Practical Guide