Guide · CRM & customer operations
What to Clean in Your CRM Before You Move It
Sort active work, duplicate contacts, unused fields, and historical records before a CRM migration turns old confusion into new confusion.
A CRM move is a good time to notice how much of the current system nobody uses. Old tags, duplicate people, and half-finished workflows don't become clearer when they land in a new database. Start with the records that support today's work, then decide what historical material you still need.
Take a copy before changing anything
Preserve an export of the original data before cleanup. Keep the original identifiers and record how the export was produced. Work from a separate copy so you can explain a change or recover a value without trying to remember what used to be there.
Treat cleanup as a sequence of decisions. Removing a blank column is different from merging two customer records. For changes that alter meaning or relationships, keep a short log of the rule, the affected records, and who reviewed it. Avoid making undocumented edits directly in your only export.
Start with open work
Ask the team which records they need to do their jobs tomorrow. Open opportunities, active customers, pending appointments, and unresolved conversations usually deserve attention before a ten-year archive. The exact list depends on your business, so use a real workday as the test.
For a hypothetical repair shop, one person can own several vehicles and several visits. A record that looks duplicated by customer name may represent distinct work. Before merging anything, identify what each row actually represents and what other records point to it.
Choose duplicate rules carefully
An identical email address can be a useful clue, but it isn't a complete identity policy. Families share addresses. Companies use shared inboxes. A missing email doesn't make a contact disposable. Inspect the fields and relationships that matter in your own system.
Create a review list for uncertain matches. Specify which record will survive, which values will be retained, and what happens to notes, opportunities, and consent information. If the migration tool doesn't support a safe merge for that data, keep the uncertainty visible and resolve it separately.
Give fields and tags a purpose
Pick a frequently used record and ask what each field helps someone decide. A custom field called "Status 2" may mean something to its creator and nothing to the next person. Rename or map it only after confirming its meaning with someone who uses it.
Group tags by purpose: a customer attribute, a campaign history marker, a temporary processing flag, or an automation trigger. They may look identical in an export but behave differently. A tag that starts an automation deserves a different migration decision from a tag used only for reporting.
Keep a short decision sheet
Before the import, record what you're moving, mapping, archiving, and rebuilding. Include anything you're leaving unresolved. This sheet makes the scope understandable to the team and gives you something specific to check after the move.
A clean import starts with agreed meanings. You can make a messy spreadsheet look tidy without solving that problem, so spend the review time on the decisions people will rely on afterward.
- The active record groups needed for current work.
- The matching rules used to identify possible duplicates.
- Field changes, with old names, new names, and example values.
- Historical material being retained outside the destination CRM.
- Automations or integrations that require separate rebuilding.