Skip to content
Tutorials Center
My learning

Topics

Lesson 10 of 37 ยท CRM Basics

HubSpot Data Quality and Deduplication

Keep HubSpot data trustworthy: find and merge duplicates, standardize values, handle missing fields, and build a routine of data hygiene before reports go wrong.

Every report, workflow and email you build trusts the data underneath. A CRM with duplicate contacts, inconsistent countries and missing owners produces confident-looking but wrong numbers. Data quality is not a one-time cleanup; it is a habit built from a few tools and a few rules.

What bad data looks like

  • Duplicates: the same person twice ([email protected] and [email protected]), or the same company as "Bluefin", "Bluefin Ltd" and "Bluefin Ltd.".
  • Inconsistent values: "USA", "U.S.", "United States".
  • Missing values: deals with no amount, no close date or no owner.
  • Stale values: job titles from five years ago, bounced emails still marked as marketing.
  • Wrong format: phone numbers with letters, names in capital letters.

Preventing duplicates

Prevention beats repair:

  1. Unique identifiers. HubSpot already refuses a second contact with the same email and a second company with the same domain through normal creation. Imports and integrations are where gaps appear, so always include the identifier.
  2. Form design. Use the email field as the key; do not make people enter a company name freehand when you can match on domain.
  3. Integration rules. When syncing from another system, decide which side wins and which key is used.
  4. Dropdowns, not free text, for values like country and industry.

Finding duplicates

HubSpot has a Manage duplicates tool (in the Contacts and Companies areas, under Actions or Data Management). It suggests pairs by comparing names, email domains and other signals, and lets you Merge or Not a duplicate. Operations Hub adds automated duplicate handling in some versions, and custom rules via data quality tools.

You can also hunt yourself:

  • Create a view sorted by Last name or Company name and scan neighbours.
  • Create a list of contacts with the same phone number.
  • Use an export and a spreadsheet pivot table on normalized names.

Merging

When you merge two records into one:

  • You choose a primary record, the one that survives.
  • Properties are combined, with the primary's values winning conflicts (you can review the choice).
  • Activities, associations and tickets move to the surviving record.
  • Contacts' emails are all kept: the extra email becomes an additional address.
  • A merge cannot be undone. Check twice, especially when two different people look alike.

Merge only when you are sure; a false merge combines two people's history and consent data.

Standardizing values

HubSpot helps you keep values uniform:

  • Data quality command center (Operations Hub Professional and above) suggests fixes such as formatting names or phone numbers, and spots inconsistent values.
  • Workflow actions can format data (capitalize, trim, change case) in Professional editions.
  • Bulk edit from a table view corrects a property for many records at once.
  • Property validation rules restrict allowed values.

A practical approach for existing mess: export, normalize in a spreadsheet or script, and import as an update keyed on Record ID or email.

Completeness: required data and ownership

Decide the minimum data each stage needs. For example: a deal cannot reach Contract sent without an amount and a close date. HubSpot's required properties on stage change (deal pipelines settings) enforce this, which stops reps from advancing deals with blank amounts.

Create views that find gaps and review them weekly:

  • Contacts with no owner.
  • Deals with a close date in the past.
  • Deals with no amount.
  • Companies with no domain.

A routine that actually happens

Cadence Task
Weekly Review the unassigned and missing-field views; resolve the duplicate suggestions.
Monthly Remove or suppress hard bounces; check unsubscribed counts.
Quarterly Review users and permissions; audit properties for unused ones.
Yearly Review lifecycle definitions and archive stale records according to policy.

Privacy law also requires you to delete or anonymize personal data you no longer need, and to honour deletion requests promptly.

Finding duplicates in code (stand-in)

The stand-ins below are not HubSpot's duplicate engine. They show two ideas behind it: normalizing before comparing, and scoring match confidence.

JS Stand-in: normalize names to find duplicate companies

Press Run (or Ctrl + Enter) to see the console output here.

No output.

JS Stand-in: scoring how likely two contacts are the same person

Press Run (or Ctrl + Enter) to see the console output here.

No output.

The second example shows why humans review merges: a score is only a hint, and the decision stays with a person.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. Which is the most effective way to reduce duplicates?

    Preventing duplicates at the source is cheaper than merging them later.
  2. 2. What is true about merging records?

    Merges are permanent, so confirm that two records are the same entity.
  3. 3. How can you stop reps advancing a deal without an amount?

    Required properties on stage change force the data to exist before the deal advances.
  4. 4. Why normalize names before looking for duplicates?

    Lower-casing, removing punctuation and legal suffixes reveals records that differ only in formatting.

Summary

  • Duplicates, inconsistent values and missing fields are the usual data problems.
  • Prevent them with unique identifiers, dropdowns and required properties.
  • Use Manage duplicates for merges, which are permanent.
  • Make data hygiene a weekly, monthly and quarterly routine.

Next: Forms and landing pages