Migrate from OpenCRM to Twenty CRM

OpenCRM Company records map directly to Twenty Company records. We extract the full OpenCRM Company dataset via CSV export including all standard fields (name, domain, address, phone) and any custom fields discovered during scoping. The Company record is imported first because OpenCRM Contacts are linked to Companies via foreign-key relationship; without the parent Company present, Contact imports would create orphaned Person records with no company association.

OpenCRM Contact records map to Twenty Person records. The OpenCRM contact_email, contact_phone, contact_firstname, and contact_lastname fields map to Twenty's Person standard fields. A key pre-migration requirement in Twenty is that many industry-standard fields (job title, department, phone type) do not exist as standard fields and must be created via the /metadata GraphQL API before Person import. We create these custom fields during the schema preparation phase so that field mapping can complete field-by-field rather than collapsing into generic text fields.

OpenCRM Deals map to Twenty Opportunities. The OpenCRM deal name, value (amount), expected close date, and owner assignment transfer directly. Pipeline stage names from OpenCRM require a stage-mapping table because Twenty Opportunity stages are tenant-defined and rarely align one-to-one with OpenCRM's workflow-specific stage names. We produce this mapping during scoping and present it for customer confirmation before the Opportunity import runs.

OpenCRM Activity records (calls, meetings, tasks) are supported for migration with timestamp normalization to UTC and owner mapping by email match to Twenty users. Twenty stores task-type activities as Task records and meeting-type activities as Event records. We validate that each activity has a valid Person or Company reference before import so the activity lands in the correct Twenty timeline.

OpenCRM Notes attached to Contacts, Companies, or Deals migrate to Twenty Comment records linked to the corresponding Person or Company. Note body content migrates as rich text. We validate parent record linkage after import to confirm notes are associated with the correct entity in Twenty.

OpenCRM custom fields on Contacts, Companies, and Deals require pre-creation in Twenty via the /metadata GraphQL API before migration begins. We discover all custom field names, data types, and object assignments during the scoping phase, then create the corresponding Twenty custom fields with matching API names. This is a required step because Twenty's Person and Company objects ship with minimal standard fields and imports that attempt to populate non-existent fields will silently fail or reject records.

Tag-based categorisation on OpenCRM Contacts and Companies migrates as flat string arrays to Twenty Tags. Tags used for content classification migrate as Twenty Tag records with TagAssignment records linking to the Person or Company. The customer confirms tag strategy during scoping since Twenty supports both object-level tags and workspace-level topic tagging.

OpenCRM user records and deal/contact ownership assignments are resolved by email match against Twenty WorkspaceMember records. Any OpenCRM Owner without a matching Twenty user goes to a reconciliation queue for the customer to provision before record import resumes. Owner resolution is required before Contacts and Deals can be imported because both record types carry an owner reference in Twenty.

OpenCRM allows full customisation of pipeline stage names per workflow. We produce a stage-mapping table during scoping that maps each OpenCRM stage name to the corresponding Twenty Pipeline stage value. The customer reviews and approves this table before Opportunity import so every Deal lands in the correct stage.

File attachments stored against OpenCRM Contacts, Companies, or Deals are extracted from the OpenCRM UI export and staged in a file store. We generate reference links in Twenty pointing to the staged file store so the attachment context is preserved even if the attachment content is not embedded in Twenty's object model. Full file embedding in Twenty is available via the /graphql API for implementations that require it.