Migrate from Pipeline CRM to Twenty CRM

Pipeline People records map directly to Twenty People. We export all standard fields (name, email, phone, address) plus any custom Person fields, then import via the Twenty GraphQL API. The twenty_person_emails primary email address maps from Pipeline's email field. We normalize phone numbers to E.164 format during the transform step. Custom Person fields in Pipeline must be pre-created in Twenty Settings before the import script runs.

Pipeline Companies map to Twenty Company records. Company name is the primary dedupe key. We import Companies first because Deals and People both reference Company as a foreign key. Industry, size, revenue, and any custom Company fields transfer directly. Pipeline's company domain (if present) maps to the Twenty Company domainName field.

Pipeline Deals map to Twenty Opportunity. The deal value (amount) and stage map from Pipeline's dealstage field to Twenty Opportunity stageName. Pipeline's pipeline assignment maps to a Twenty Opportunity record type or stage group that we configure before migration. Closed dates and probability percentages transfer as Opportunity closeDate and probability. We resolve the Company reference on each Deal by matching on Pipeline Company name.

Each Pipeline CRM pipeline maps to a Twenty Opportunity stage group. Pipeline stages are created as stageName values in Twenty with corresponding probability percentages. If Pipeline uses multiple pipelines (up to 20 on Develop, unlimited on Grow), we create multiple stage groups in Twenty and assign each Deal to the appropriate group via the record's stageGroup field.

Pipeline Activities (email, call, meeting history linked to People or Companies) map to Twenty Task records. Activity type (email, call, meeting) maps to Task category or type fields in Twenty. We preserve the activity timestamp as Task dueDate for ordering in the timeline. The linked Person or Company reference is resolved using the Pipeline relationship keys present in the export CSV.

Pipeline Agenda Tasks migrate to Twenty Task records with Status, Priority, and dueDate preserved. Task assignee resolves by matching Pipeline owner email to a Twenty user that has been provisioned before this import phase begins. Completed status migrates as Task completedAt.

Pipeline Agenda Events (calendar items) migrate to Twenty records. If Twenty has a native Event object, Events map directly with StartDateTime, EndDateTime, and Location preserved. If the deployed Twenty version does not include Events, Events migrate as Task records with type=Meeting and the meeting window stored in the description or a custom field.

Pipeline tags applied to People, Companies, and Deals migrate as label arrays or comma-separated strings in Twenty custom fields. We ask the customer during scoping whether they want tags as a native Twenty label field or as a multi-select custom field on each object. Multi-select fields must be pre-created in Twenty Settings before import.

Pipeline Owners assigned to Deals and People export by name and email. We match each Pipeline Owner email to a Twenty user that the customer's admin has provisioned in Settings before migration begins. Any Pipeline Owner without a matching Twenty user is held in a reconciliation queue. We do not create Twenty users automatically; that requires an explicit admin invitation in Twenty Settings.

Pipeline custom fields defined per object (Company, Deal, Person) export alongside the standard fields in CSV. Twenty custom fields must be pre-created in Settings before the import script runs because the Twenty CSV import creates records, not fields. We export the full Pipeline field schema during discovery, map field types to their closest Twenty equivalents (text, number, date, checkbox, select), and document any fields that cannot map directly and require transformation.

Pipeline file attachments linked to People, Companies, or Deals export with a URL reference. We preserve the attachment URL in a custom field on the target Twenty record and flag that file hosting must be managed separately in Twenty's storage (which varies by self-hosted vs cloud deployment). Native attachment upload into Twenty requires separate handling beyond the CSV import path.