Migrate from Sercom to Twenty CRM

Sercom Contact maps to Twenty's People object. The People object stores name, email, phone, job title, and a companyId relation. Sercom contacts without a primary company land in Twenty People with no companyId — that relation is resolved during the import pass once all Company records exist.

Sercom Account maps to Twenty's Companies object. Companies must be imported before People because People records carry a companyId foreign key. FlitStack AI sequences the import so all Accounts resolve to Twenty Companies before any Contact-to-People record is written. This ordering ensures that every People record receives a valid companyId reference rather than a null or broken link.

Sercom Deal maps to Twenty's Opportunities object. Opportunities store deal name, amount, stage, probability, expectedCloseDate, and linked companyId and personId relations. Opportunities import third in sequence after Companies and People because they carry two foreign keys that must resolve to already-existing records.

Sercom Task maps to Twenty's Tasks object. Tasks carry a name, status, dueDate, completedAt timestamp, and a workspaceMemberId relation for the assignee. Tasks can be linked to any record (People, Companies, Opportunities) via Twenty's generic relation model. Original create and update timestamps are preserved as custom fields since Twenty's Tasks object does not expose custom datetime fields natively.

Sercom Note maps to Twenty's Notes object. Notes store a body (rich-text content) and can be attached to People, Companies, Opportunities, or other records. FlitStack AI preserves the original createdAt timestamp as a custom field and maintains the parent-record link so note context is not lost during the migration.

Sercom's custom fields on Contact (such as source_system_id, original_create_date, or any industry-specific pick-list) require pre-creation in Twenty's Settings → Data Model before the CSV import. FlitStack AI generates a field-creation checklist based on Sercom's custom field inventory so Twenty's schema is ready before data lands.

Sercom's custom fields on Account migrate to Twenty Companies custom fields. Fields like annual_revenue, number_of_employees, and industry (where not matched to Twenty's standard pick-list) are created in Twenty Settings → Data Model first. Custom pick-list options are preserved as written values rather than enforced enums since Twenty's select fields are configurable post-import.

Sercom's custom fields on Deal (deal_priority, probability, custom revenue fields) migrate as Twenty Opportunities custom fields. The field type is carried over (number, text, select) and pre-created in Settings → Data Model. Probability and deal_priority are stored as custom number and select fields respectively since Twenty Opportunities does not expose these natively.

Sercom custom objects map 1:1 to Twenty custom objects. Custom object definitions (name, field list, field types) are extracted from Sercom and translated into Twenty's metadata API format. Custom object associations that use a many-to-many model in Sercom need a junction object in Twenty — FlitStack AI surfaces this in the pre-migration plan.

Sercom file attachments linked to records are preserved as external URLs in a custom text field (File_URL__c) pointing to the original storage location, or re-uploaded to Twenty's connected storage. Files that are inline in notes are extracted, downloaded, and stored as attachments in Twenty's file management system.

Sercom owner records (sales reps assigned to contacts, deals, and tasks) resolve by email match against Twenty WorkspaceMembers. Unmatched owners are flagged before migration — the team either creates the Twenty user first or assigns records to a fallback WorkspaceMember. Owner resolution is required before the People import pass runs.

Sercom's call logs, meeting records, and email history (where stored as activity objects rather than notes) migrate as Twenty Tasks with the activity type preserved in a custom field (Activity_Type__c). Original timestamps and assignee information are carried forward. This assumes Sercom exposes these activity records via its export or API.