Migrate from PBS Systems to Twenty CRM

Audit PBS data and build the migration mapping document

FlitStack AI begins every PBS migration with a data audit using the PBS database export interface or partner API channel. We inventory all customer records, deal records, vehicle records, activity logs, service write-ups, and custom fields present in the PBS export. We then build a detailed mapping document that specifies for every field: the PBS field name, the Twenty object and field name, the mapping type (direct, transformed, value-mapped, custom field required), and any transformation logic required. This document is reviewed with your team before any field creation or data movement begins. The audit also identifies records that will not migrate cleanly — orphaned records, records with invalid foreign keys, and PBS-specific data with no Twenty equivalent — so these can be resolved or documented as intentional exclusions.

Create Twenty custom fields and custom objects before import

Before any CSV data is uploaded to Twenty, FlitStack AI creates all necessary custom fields and custom objects in the Twenty workspace via Settings → Data Model. This includes the Vehicle custom object with VIN__c, Make__c, Model__c, Year__c, Color__c, Mileage__c, and StockNumber__c fields; custom fields on the Opportunity record for APR__c, Finance_Term__c, Down_Payment__c, Trade_In_Value__c, and FI_Products__c; custom fields on People for Original_Create_Date__c and Source_System_ID__c; and the ServiceWriteUp custom object with labor and parts fields. Custom pick-list values are populated with the exact stage names and F&I product names from the PBS export so value-mapping is accurate. This step ensures that Twenty's import validator does not reject any CSV rows for missing field references when the migration runs.

Resolve PBS owners and users by email against Twenty Workspace Members

PBS owner IDs and user records are resolved by email match against the Twenty workspace member list. FlitStack AI generates a pre-migration owner resolution report that lists every PBS owner, their email, their Twenty match status (matched, not found, or duplicate email), and the proposed resolution (map to matched user, assign to fallback owner, or flag for admin decision). PBS users who do not yet have a Twenty account are either invited before migration or their records are assigned to a designated fallback Twenty user so no migrated record lands without an owner. This step is critical because Twenty's Task and Opportunity records require an assigneeId relation to a Workspace Member, and unresolvable owners would cause those records to fail import validation.

Sequence and run the migration: Companies → People → Opportunities → Custom Objects → Activities

FlitStack AI sequences the migration to respect Twenty's foreign-key constraints. Companies are migrated first (the 'one' side of the relationship), then People with their companyId relations resolved to the migrated Company records, then Opportunities with companyId and personId relations resolved, then Vehicle and ServiceWriteUp custom objects linked to their parent Opportunity or People records, and finally Tasks and Notes attached to the relevant records. Each stage runs a validation pass that checks record counts, required field completeness, and foreign-key resolution before the next stage begins. Original create dates, modified timestamps, and owner assignments are preserved throughout. A field-level diff is generated after each stage comparing source values against the Twenty record to verify mapping accuracy.

Cut over with delta-pickup and one-click rollback available

The full migration runs against Twenty's API with a delta-pickup window of 24–48 hours after the primary load completes. During this window, any PBS records created or modified after the initial extraction are captured and applied to Twenty. Your team continues working in PBS during the cutover period — FlitStack AI uses scoped read access only, so no PBS data is altered. An audit log records every operation performed in Twenty (insert, update, skip). If reconciliation fails — a field mapping produces incorrect values, a record set is incomplete, or a custom object fails to link — FlitStack AI executes a one-click rollback that reverts the Twenty workspace to its pre-migration state, allowing the migration to be re-run with corrected mapping after the issue is resolved. Post-migration, a final validation report is delivered comparing PBS record counts against Twenty record counts with a field-by-field completeness score.