How to Run a HubSpot Data Quality Audit in 2026

Your HubSpot portal is supposed to be the operational backbone of your sales and marketing team. But when was the last time you actually audited the quality of that data?

Poor data quality is expensive — Gartner's research consistently puts the cost at millions per year for mid-to-large organisations. In HubSpot specifically, bad data leads to failed automations, inaccurate reporting, wasted sales time, and missed opportunities. Yet most companies only discover their data problems when something breaks.

This guide walks through how to run a full HubSpot data quality audit in 2026, whether you're doing it manually or using automated tools. The 10 audit areas below map to the six formal data quality dimensions — accuracy, completeness, consistency, validity, uniqueness, and timeliness — so you have a vocabulary for what you're measuring and why.

When should you run a data quality audit?

Most HubSpot admins should audit their data quality in these scenarios:

Quarterly maintenance cycles. Treat data audits like oil changes for your CRM. Even well-maintained portals accumulate cruft over time as properties are created, workflows evolve, and teams change.

Before major initiatives. Planning a migration, implementing AI tools, or launching a new revenue operations strategy? Audit first. Building on bad data amplifies problems exponentially.

After significant growth. If your contact database has doubled, your team has expanded, or you've acquired another company, your data structure likely needs attention.

When reporting feels wrong. If stakeholders question your dashboards or sales reps complain about data accuracy, those are symptoms of deeper issues.

Post-integration chaos. Connected a new tool to HubSpot? Integrations often import messy data or create duplicate fields. Audit within 30 days of major integration changes.

The best approach? Schedule audits quarterly and run spot-checks whenever something feels off. Prevention is cheaper than remediation. Not sure what a healthy CRM looks like? Our guide on CRM health scores explains the benchmarks and what each score component measures.

The 10-area HubSpot data quality audit checklist

A thorough audit covers more than just "are there duplicates?" Here are the ten critical areas every audit should examine:

1. Property naming conventions

Inconsistent property names are the gateway drug to data chaos. Over time, different team members create properties with overlapping purposes but different labels.

What to check:

• Do property names follow a consistent pattern (snake_case, camelCase, or Title Case)?
• Are abbreviations standardized (e.g., "phone" vs "ph" vs "tel")?
• Can you identify a property's purpose from its name alone?
• Are there duplicate properties with slightly different names (lead_source vs LeadSource vs Source)?

Red flags: Properties named "test," "temp," "old," or date-stamped versions (lead_score_2024, lead_score_2025). These indicate cleanup debt.

Best practice: Establish a naming convention document and enforce it. Modern portals often use prefixes to indicate custom properties (custom_lead_score) or departmental ownership (sales_priority_score).

2. Duplicate records

Data quality dimension: uniqueness

Duplicate Contacts, Companies, and Deals fragment your data and distort reporting. They're especially common after imports, integrations, or when multiple team members manually add records.

What to check:

• Run HubSpot's native duplicate detection (Contacts > Actions > Manage duplicates)
• Look for near-duplicates: same company, different domains; same person, different email addresses
• Check for role-based duplicates (info@company.com, support@company.com both mapped to Company X)
• Review recently created records for import-related duplicates

Red flags: Spike in duplicate rates after imports, contacts at the same company with wildly different lifecycle stages, or multiple deals for the same opportunity.

Best practice: HubSpot's automatic duplicate management helps, but it's not perfect. Configure merge rules, train teams on search-before-create, and clean up systematically.

3. Property fill rates

Data quality dimension: completeness

Empty fields are useless fields. Low fill rates indicate properties that aren't being used, poorly designed forms, or gaps in data collection processes.

What to check:

• Calculate fill rate (percentage of records with data) for each custom property
• Identify properties below 20% fill rate—these are candidates for deletion
• Check required fields on forms vs actual completion rates
• Review high-value properties (lead source, industry, company size) for gaps

Red flags: Properties with <5% fill rate that are used in workflows or reports, or properties that were required in the past but aren't collecting new data.

Best practice: Archive or delete unused properties. For important low-fill properties, identify why they're not populating and fix the process (better forms, integration mapping, data enrichment).

4. Inactive and obsolete properties

HubSpot portals accumulate properties like attics accumulate boxes. That "temporary" field from 2021? Still there. The property from a discontinued tool? Haunting your settings.

What to check:

• Review properties created more than 18 months ago that haven't been updated
• Check properties associated with deprecated integrations
• Look for properties that are no longer used in any workflows, reports, or lists
• Identify seasonal or campaign-specific properties that are now obsolete

Red flags: Properties referencing old campaigns, discontinued products, or former team members. Field descriptions that say "don't use this anymore."

Best practice: Archive properties rather than deleting (preserves historical data). Our guide on finding and removing unused properties covers the 90-day phased cleanup process. Use property descriptions to mark deprecation dates. Schedule annual "property purge" reviews.

5. Data type mismatches

Data quality dimension: validity — whether data conforms to defined formats and rules

Using text fields for numbers, number fields for dates, or single-line text for paragraphs creates problems for reporting, segmentation, and integrations.

What to check:

• Properties storing numbers as text (prevents math operations)
• Date information stored as text (breaks date-based automation)
• Dropdown values when radio select or checkboxes would be better
• Multi-line text fields containing single words
• Enumerated options stored as text ("true"/"false" instead of true boolean)

Red flags: Reports that can't calculate sums or averages, automation rules that fail unexpectedly, or CRM records where sorting doesn't work logically.

Best practice: Plan data types before creating properties. Understand that some type changes are destructive—HubSpot may clear data when converting between incompatible types.

6. Required field strategy

Required fields enforce data quality at the entry point, but too many requirements frustrate users and reduce conversion rates.

What to check:

• Which properties are marked as required on forms?
• Are required fields actually necessary for your processes?
• Do required fields differ by form/context appropriately?
• Are sales reps bypassing requirements using dummy data?

Red flags: Forms with 10+ required fields, low form conversion rates, obvious fake data (test@test.com, 555-5555), or identical values across multiple records.

Best practice: Require only what's needed for immediate next actions. Use conditional logic to require different fields based on context. Progressively profile—collect more data over time rather than upfront.

7. Association mapping

HubSpot's power comes from connecting Contacts to Companies to Deals to Tickets. Broken associations break that power.

What to check:

• Contacts without associated Companies (especially for B2B businesses)
• Deals without associated Contacts or Companies
• Companies with no Contacts
• Tickets missing Contact associations
• Multiple contacts at the same company with conflicting information

Red flags: Large numbers of unassociated records, especially after imports. B2B Deals assigned to personal email domains. Companies with 50+ associated contacts (might indicate incorrect association rules).

Best practice: Use HubSpot's association settings to automate contact-to-company matching. Our guide on fixing broken associations covers bulk-fix techniques and prevention governance. Review association rules after imports. Train teams on manual association best practices.

8. Lifecycle stage integrity

Data quality dimension: consistency — do related values align logically across records and objects?

Lifecycle stages drive workflow logic, reporting, and lead routing. Inconsistent lifecycle data causes contacts to skip stages, regress illogically, or get stuck.

What to check:

• Contacts that jumped stages (MQL to Customer without SQL/Opportunity)
• Contacts moving backward through stages
• Deal stages that don't match contact lifecycle stages
• Orphaned lifecycle stages no longer used in processes
• Default lifecycle stage assignments on import

Red flags: Marketing Qualified Leads with closed-won deals, Customers marked as Subscribers, or large percentages of records in "Other" catch-all stages.

Best practice: Document your lifecycle stage definitions. Use workflow automation to enforce progression logic. Regularly review stage transition reports to catch anomalies.

9. Lead scoring accuracy

Lead scores should correlate with actual conversion rates. If high-scoring leads don't convert or low-scoring leads become customers, your scoring model is broken.

What to check:

• Distribution of lead scores across your database
• Conversion rates by score bracket
• Properties contributing to score calculations
• Score decay rules for stale leads
• Negative scoring for disqualifying behaviors

Red flags: All leads clustering around the same score, scores that never decrease, or high-scoring leads that sales ignores.

Best practice: Calibrate scoring quarterly based on actual conversion data. Use both demographic (fit) and behavioral (engagement) signals. Implement score decay for aging leads.

10. Reporting property health

Properties used in reports and dashboards require higher quality standards. Missing data here directly impacts decision-making.

What to check:

• Identify properties used in business-critical reports
• Calculate fill rates for these high-value properties
• Review data consistency (same information formatted differently)
• Check for properties used in formulas that have null values
• Validate properties used in forecasting or revenue operations

Red flags: Executive dashboards with "no data available," inconsistent formatting (US dates vs EU dates), or reports that exclude large percentages of records due to missing property values.

Best practice: Tag reporting properties in your documentation. Set higher data quality thresholds (90%+ fill rate). Implement validation rules to enforce data entry standards.

How audit areas map to data quality dimensions

Each of the 10 audit areas above measures one or more of the six formal data quality dimensions. This mapping helps you prioritise — if a specific dimension scores poorly, you know exactly which audit areas to focus on:

Audit Area	Primary Dimension	Secondary Dimension
Property naming conventions	Validity	Governance
Duplicate records	Uniqueness	—
Property fill rates	Completeness	—
Inactive/obsolete properties	Timeliness	Governance
Data type mismatches	Validity	Accuracy
Required field strategy	Completeness	Governance
Association mapping	Consistency	Accuracy
Lifecycle stage integrity	Consistency	Validity
Lead scoring accuracy	Accuracy	Timeliness
Reporting property health	Completeness	Accuracy

Most audit areas map to a single primary dimension, though governance underpins several. If you're new to the formal dimensions framework, our data quality dimensions guide explains each one with HubSpot-specific examples and benchmarks.

Manual audit process: Step-by-step

If you're running your audit manually, here's the systematic approach:

Week 1: Property Inventory Export your entire property list (Settings > Properties). Create a spreadsheet tracking: property name, type, group, created date, last modified, used in workflows (Y/N), used in reports (Y/N), fill rate.

Calculate fill rates using HubSpot's reporting tools or by exporting contact/company samples. Flag properties below 20% for review.

Week 2: Data Exports and Analysis Export records from each object type (Contacts, Companies, Deals, Tickets). Use spreadsheet formulas to identify:

• Duplicates (VLOOKUP for matching emails, domains, deal names)
• Data type issues (ISNUMBER tests for fields that should contain numbers)
• Association gaps (contacts without company IDs)
• Lifecycle anomalies (stage progression logic)

Week 3: Workflow and Report Review Audit every active workflow. Document which properties each workflow uses. Note workflows touching deprecated properties or using complex logic that might break.

Review all business-critical reports and dashboards. Document property dependencies. Test report accuracy with sample data.

Week 4: Documentation and Cleanup Create a findings document with prioritized issues:

• Critical (breaks workflows, blocks deals)
• Important (impacts reporting accuracy)
• Maintenance (cleanup for efficiency)

Execute cleanup systematically, testing after each major change. Update your data governance documentation with new standards.

Time investment: 40-60 hours for a portal with 100-200 custom properties and 50K+ records.

Automated audit approach: Modern tools

Manual audits work, but they're time-intensive and error-prone. Automated tools can analyze your entire portal in minutes rather than weeks.

HubSpot native tools

HubSpot provides basic audit capabilities:

• Property settings show "last used" dates
• Duplicate management tool identifies matching records
• Reports can calculate basic property fill rates
• Association settings show mapping rules

Limitations: No overall health scoring, no cross-object analysis, no prioritized recommendations, no historical tracking.

Third-party automation

Tools like HubHorizon analyze your entire portal structure automatically, checking:

Property health scoring: Automated fill rate calculation, usage tracking, naming convention validation, and obsolescence detection across all object types simultaneously.

Relationship mapping: Automated association analysis identifying orphaned records, bi-directional relationship validation, and company-contact matching accuracy.

Data consistency checks: Cross-property validation (do email addresses match email format?), lifecycle logic verification, and data type conformance testing.

AI readiness assessment: Analysis of how well your data structure supports modern AI tools, including property documentation completeness, data standardization, and semantic consistency.

Prioritized recommendations: Instead of a massive list of issues, automated tools rank problems by business impact—fixing broken workflow dependencies before renaming cosmetic property issues.

Time investment: Initial setup (30 minutes), automated analysis (5 minutes), review and action planning (2-4 hours).

The difference is speed and coverage: automated tools identify issues you'd miss manually, quantify impact you'd have to estimate, and reduce a month-long project to an afternoon.

What to do with audit results

Discovering problems is only valuable if you fix them. Here's how to prioritize remediation:

Tier 1 (Fix Immediately):

• Properties breaking active workflows
• Duplicate records confusing sales teams
• Missing associations blocking deal progression
• Data type issues preventing calculations

Tier 2 (Fix This Quarter):

• Low fill rates on reporting properties
• Lifecycle stage inconsistencies
• Inactive properties cluttering settings
• Lead scoring miscalibrations

Tier 3 (Ongoing Maintenance):

• Naming convention standardization
• Progressive profiling implementation
• Documentation improvements
• Training and process updates

Create a Kanban board or project in your work management tool. Assign owners. Set deadlines. Track completion.

Most importantly: fix the process, not just the data. If your audit revealed 50 duplicate contacts from a recent import, deleting them solves nothing if the next import creates 50 more. Update your import mapping, train the responsible team member, and document the correct process.

Building a data quality culture

One-time audits help, but lasting data quality requires cultural change:

Governance documentation: Maintain a living document covering naming conventions, property creation approval process, required vs optional field strategy, and lifecycle stage definitions. Our data governance policy guide provides a ready-made template.

Onboarding training: Every new team member touching HubSpot should learn data quality standards before they create their first property or import their first list.

Regular reviews: Schedule quarterly audits. Make data quality a standing agenda item in RevOps meetings.

Measurement and accountability: Track data quality metrics (duplicate rate, property fill rates, workflow error rates) like you track revenue metrics. Share results with stakeholders.

Tools and automation: Implement validation rules, required fields, and workflow-based data cleaning. Use automation to enforce standards rather than relying on human discipline.

Perfect data is impossible. Disciplined data that supports business decisions and scales with growth — that's the target.

Frequently Asked Questions

How often should you audit your HubSpot data?

Quarterly is the recommended minimum for a formal data quality audit. Monthly automated checks catch degradation between full audits. You should also trigger an audit after major events: bulk imports, new integrations going live, CRM migrations, significant team changes, or before implementing AI tools like HubSpot Breeze.

What does a CRM data quality audit check?

A comprehensive HubSpot audit covers 10 areas: property hygiene, data completeness, duplicate records, association health, naming conventions, lifecycle stage accuracy, workflow dependencies, reporting reliability, integration data flows, and permission/governance structure. Each area has measurable indicators — fill rates, duplicate counts, naming compliance scores — that produce an overall health score.

How long does a HubSpot data quality audit take?

A manual audit of a mid-size portal (10,000-50,000 records) typically takes 2-5 days of focused work across the 10 key areas. Automated tools like HubHorizon reduce this to under 5 minutes for the diagnostic phase, letting you spend your time on remediation instead of discovery.

Can you automate a HubSpot data quality audit?

Yes, most diagnostic components can be automated: property analysis, fill rate calculation, duplicate detection, association health checks, and naming convention compliance. What can't be fully automated is the interpretation — understanding which issues matter most for your specific business context — and the remediation itself. The best approach combines automated diagnostics with human prioritisation.

Start your automated audit

Data quality isn't a one-time project; it's an ongoing discipline. But every discipline starts with assessment.

Whether you choose manual analysis or automated tools, running a full HubSpot data quality audit gives you clarity on what's broken, a roadmap to fix it, and confidence in the data driving your business.

For teams ready to move beyond manual spreadsheet analysis, HubHorizon automates the audit process. Connect your portal, receive a full analysis in minutes, and get prioritized recommendations for improvement.

Start your automated HubSpot data quality audit now →

Your reporting gets more accurate, your automation stops breaking, and your sales team stops complaining about garbage data. View pricing plans for continuous monitoring, AI readiness scoring, and quarterly audit reports.