CRM Data:
What It Includes and How to Keep It Current

CRM data is the information stored in a customer relationship management system about contacts, customers, and their interactions with a business. It includes contact details, communication history, purchases, service requests, and preferences, according to Salesforce’s CRM database guide. Managing that data involves choosing useful fields, linking records correctly, identifying their sources, and keeping changing information current.

What types of data does a CRM contain?#

CRM data covers both customer information and the history of the relationship. The following categories describe common contents; select fields according to the work they need to support.

How CRM data is organized#

The basic distinction is between a record, its fields, and its relationships. In HubSpot’s data model, an object is a category of information, a record is one instance of that object, a property is a field on a record, and an association links records. Contacts, companies, deals, and tickets are separate objects.

Use the same distinctions when designing data in another CRM:

• Store person-level information on the contact record.
• Store organization-level information on the company or account record.
• Store transaction details on the relevant deal, order, or purchase record.
• Store support details on the relevant request or ticket.
• Check the links between those records as part of data maintenance.

Keep dated interactions separate from current summary fields. Preserve a communication’s date and content in the activity history; define an update rule for any summary derived from that history. Review associations during imports and cleanup, alongside the values in individual fields.

Where CRM data comes from#

A record’s creation source explains how it entered the CRM. HubSpot’s Record Source documentation identifies creation methods such as imports, forms, and workflows; its source details can identify the specific import or form. The documentation distinguishes these properties from Original Traffic Source and Latest Traffic Source, which describe interactions with web content.

Keep these questions separate:

• Record origin: Which process created the record?
• Acquisition source: Which channel brought the contact to the business?
• Field update source: Which process or person changed a particular value?

For field changes, HubSpot’s property history view shows previous values, change dates, and sources. Check that history when investigating an unexpected update.

For each important field, document which sources may write to it and how conflicts should be resolved. Specify whether a confirmed correction takes precedence over an imported value, and identify who reviews unresolved conflicts. Keep the creation source and the latest field update source available for different purposes.

How to choose CRM data fields#

Begin with the work the record must support: contacting someone, progressing a transaction, resolving a request, or producing a defined report. List the inputs needed for that work before adding custom fields.

Create a short data dictionary for each retained field:

Configure the software to reflect that dictionary. HubSpot’s property editor supports field types, selection options, descriptions, entry instructions, and validation rules, with available settings varying by field type and account configuration.

Use a date field for a date, a numeric field for a quantity, and defined choices for a controlled status. Reserve free text for information that needs an explanation. Decide when a required value can be supplied, and preserve a defined unknown state where appropriate.

Assign responsibility for definitions and corrections separately from technical administration. Name who confirms the fact, who maintains the field configuration, and who investigates failed updates.

How to collect and import CRM data#

Review the destination fields and record relationships before loading a file. HubSpot’s import formatting documentation requires column headers corresponding to CRM properties and specifies formatting requirements by field type. Updates to existing records require a unique identifier, such as Record ID, email, or company domain name, depending on the object.

Prepare an import mapping that records:

1. The source column and destination field.
2. The destination object and record identifier.
3. Any conversion of dates, categories, numbers, or currencies.
4. The treatment of missing values and conflicting values.
5. The associations that need to be preserved.

Test a limited import and inspect the resulting records, values, and links before processing the full file. Check the error output as well as the records that loaded.

Pay particular attention to blanks: HubSpot’s import tool ignores blank spreadsheet cells, so a blank cell does not clear an existing property value. Check the corresponding behavior in the destination CRM.

How to keep CRM data current#

Use field-level update rules. Record the event that should change a value, the responsible person or source, and the maximum delay acceptable for its use. Also define what happens when an update is missed: a review task, an exception flag, or exclusion from a report that requires confirmed data.

The following are recommended maintenance rules to adapt to the records being managed:

Track the time a fact was confirmed separately from the time software last wrote to the record when that distinction matters. HubSpot’s Last modified date property reflects an update to any contact property, including internal properties that may be hidden. It does not identify which visible field was confirmed.

Review the relevant field history to see what changed, when, and through which source. Treat that information as evidence of an update; confirm the underlying fact through the appropriate source before treating it as verified customer information.

How to measure CRM data quality#

The UK Government Data Quality Framework defines quality in relation to intended use and describes six dimensions. These provide a useful vocabulary for CRM checks:

For a field completion measure, divide records with the required value by records that need it. State the field, eligible records, and reporting period. Check accuracy separately: a populated field can contain an incorrect value, as the framework explains.

Assign each failed check a correction owner. Prioritize fields used for active communication, unresolved requests, current transactions, or a report being used to make a decision.

How to clean CRM data#

Use a defined sequence for reviewing and correcting records:

1. Identify the records and fields in scope, and preserve the information needed to review changes.
2. Flag missing required values, invalid formats, conflicting categories, and broken associations.
3. Review potential duplicates against identifying information and related history.
4. Confirm the correct values through the designated source or record owner.
5. Apply corrections and recheck the affected workflows and reports.
6. Adjust the entry process, import mapping, or integration responsible for repeat errors.

Check merge behavior before combining records. HubSpot’s merge documentation says the remaining record combines activities, associations, and most property values, and that records cannot be unmerged. Review the values and relationships to retain before approving a merge.

Use built-in tools where their checks match the work. HubSpot’s data quality tools surface duplicates, formatting issues, enrichment coverage, and property usage; availability and actions depend on subscriptions and permissions. The documentation also notes that a formatting problem corrected through its overview may not be flagged again after a later incorrect manual update. Keep entry controls and follow-up checks in the maintenance process.

How CRM data is used in workflows and reports#

CRM data supports customer segmentation, sales opportunity tracking, and analysis of service inquiries, as outlined in Salesforce’s explanation of data use.

For each report, define the record being counted, the qualifying condition, the reporting period, and the date used to place a record in that period. Distinguish a contact creation date from a purchase date or a transaction milestone. Document how duplicates, unknown values, and excluded records are handled.

Trace reported totals back to the underlying records and inspect their relationships. Review both included and excluded records against the report’s definition. HubSpot’s report data quality insights can display property issues and fill rates; use those measures alongside checks of the facts and associations that the report needs.

Managing access to CRM data#

Decide who needs to view records, edit fields, import information, and export it. HubSpot’s user permissions guide describes record-level view and edit permissions and separate controls for importing and exporting CRM data.

Assign access according to the work being performed. Review permissions when responsibilities change, restrict broad edits and exports to the people who need them, and keep customer information out of unrelated notes and unrestricted shared files. Include connected applications in the review of who can update important fields.

Frequently asked questions#