TToolPickly
ToolPickly Editorial Research

CRM migration checklist: preserve associations before leaving HubSpot or Pipedrive

Design a migration drill around contacts, companies, deals, activities, files, and associations instead of treating one contact CSV as a complete CRM backup.

10 minute readHow we work
Decision in one minute

Before an annual contract, run a two-way migration drill with 100–300 real contacts, 20 deals, and activity history. The exit path exists only if fields, owners, stages, associations, and history can all be restored.

01 | Split CRM data into six asset classes

Contacts are only one part of a CRM. Inventory contacts, companies, deals or leads, activities and notes, files, and object associations separately. Preserve custom-field definitions, pipeline stages, owners, and enum values as well, or the records may import without restoring the operating model.

HubSpot documents that record exports can include current property values and associations while activities require separate exports; contact exports can optionally include property history. Pipedrive likewise directs users to export activities, notes, and files separately. In both cases, one contact CSV is not a complete backup.

  • Are field names, types, required rules, and option values preserved?
  • Can contact, company, and deal associations be rebuilt?
  • Do activities, notes, files, and ownership have separate exports?

02 | Define field mapping and deduplication before import

Create a mapping sheet with source field, destination field, data type, sample value, transformation rule, and null handling. Preserve email, company domain, and source-system IDs as candidate unique identifiers. Test dropdowns, dates, currencies, and multi-select fields separately rather than deciding mappings during the first import.

Pipedrive documents spreadsheet imports for people, organizations, leads, deals, activities, notes, products, and other objects, with duplicate behavior based on identifying fields. Before migration, define which records merge, which are created, and which system wins when values conflict.

03 | Run a reversible migration drill with a small sample

Choose 100–300 contacts covering normal and edge cases, 20 deals across stages, 30 days of activities, and several files. Import into a test environment, then reconcile counts, fields, owners, stages, associations, timestamps, and links. Do not begin with the full production dataset.

Put the rollback window in the plan. Pipedrive currently documents a 48-hour import reversion window. Whether or not the destination provides rollback, retain raw exports, transformed files, import logs, and skipped records until the business owner signs off.

04 | Validate workflows after cutover, not only record counts

Migration success is more than imported row count. Have sales create a lead, advance a stage, send an email, log a meeting, assign an owner, and run a core report. Marketing and service teams should verify lists, subscription state, consent records, and service history.

Keep at least one week of parallel reconciliation and freeze unnecessary field or automation changes. After cutover, document unmigrated objects, manual remediation hours, the old system's read-only period, final deletion date, and owner. All are part of true migration cost.

Official sources and evidence

Product behavior and search-report definitions can change. Recheck the linked official pages before a purchase or migration.