Migrate from Spark CRM to Twenty CRM

Spark contacts map 1:1 to Twenty's People object. The contact's name, email, phone, job title, and custom properties migrate directly. The company association requires a companyId lookup — Spark contacts without a linked company land in Twenty with no companyId and can be linked manually or via a post-migration script.

Spark companies map to Twenty's Companies object. Company name, domain/website, industry, employee count, and annual revenue migrate directly. Spark company hierarchies (parent/child) map to Twenty's Parent CompanyId field if present. Multi-company associations on a single contact collapse to one primary companyId in Twenty — the most recently modified or primary-flagged company becomes the linked record.

Spark deals map to Twenty's Opportunities object. Deal name becomes the Opportunity name, deal amount maps to amount, and close date maps to closeDate. Spark deal stage names require value mapping to Twenty's opportunity stage pick-list — we capture the mapping in the migration plan before data lands. Deals without a linked company or contact attach to the Opportunity record without a foreign key; post-migration linking is available.

Spark notes migrate as Twenty notes. The note body maps to the Body field, and original create/update timestamps are preserved as custom datetime fields (CreatedAt__c and UpdatedAt__c) since Twenty's native timestamp reflects migration time. Notes linked to contacts or companies retain their parent record association via the Note's relationId field.

Spark tasks migrate as Twenty tasks. Task title becomes the Subject field, due date maps to dueDate, and completion status maps to status (completed/incomplete). Assigned owner resolves to a Twenty WorkspaceMember by email match. Tasks linked to contacts, companies, or deals carry their relationId forward so the task appears on the correct record in Twenty.

Spark's owner field (user/rep assigned to a contact, company, or deal) resolves to Twenty WorkspaceMembers by email address. Unmatched owners are flagged before migration — your team either creates the Twenty user first or assigns records to a default owner. This step runs before any data migration so every record lands with a valid owner reference.

Spark custom fields on contacts require Twenty custom field creation via the Metadata API before migration begins. We create the field in Twenty matching the Spark field type (text, number, date, picklist) and then migrate values row-by-row. Picklist fields in Spark require value-by-value mapping to Twenty's select field options — we document this mapping in the pre-migration plan.

Spark custom fields on deals migrate as Twenty custom fields on Opportunities. The same Metadata API field creation process applies. For numeric custom fields (deal score, probability override), we preserve the raw value as a Twenty number field. Date custom fields map to Twenty date fields with original dates intact.

Spark's tag field (contact tags, company tags) requires transformation into Twenty's multi-select field type. We create a custom multi-select field called Tags__c on the People or Companies object and populate it with Spark's comma-separated tag values. If Spark has multiple tag fields, each becomes a separate multi-select field in Twenty.

Spark file attachments are not included in CSV exports and cannot be migrated programmatically via Twenty's current import API. We document the attachment list (record + filename) in a reference CSV so your team can manually re-upload files to the correct Twenty records, or use Twenty's API to upload files post-migration if the attachment volume is manageable.

Spark deal stages map to Twenty's opportunity stage pick-list via value-by-value mapping. We capture every unique Spark stage name, map it to a Twenty stage with matching semantics (e.g., 'Closed Won' → 'Won'), and apply the mapping during the Opportunity migration. Stage order and probability mapping are documented for your Twenty admin to configure post-migration.

Spark's activity history (logged calls, emails, meetings) migrates as a combination of Twenty tasks and notes. Call logs become tasks with Type='Call' and the call outcome in the task body. Email subjects become task subjects. Meeting records become tasks with a note attachment containing meeting details. Original activity timestamps are preserved in custom datetime fields.