Migrate from Thread to Freshdesk

Thread review records (including source platform, star rating, review text, and response status) map to Freshdesk Ticket records. We create custom fields on the Freshdesk Ticket object to store the original review source URL, star rating, and response timestamp. The review text becomes the ticket description. This mapping requires the Freshdesk API and therefore a paid tier (Blossom or above); the Sprout free tier does not expose the API needed for automated import.

Thread review responses from agents map to Freshdesk Ticket conversations with the type set to reply. We enforce a parent-record dependency: review responses are queued for import only after their parent review record has been successfully written to Freshdesk as a Ticket. Without this sequencing, responses become unlinked text in Freshdesk with no connection to the original review record.

Thread conversations are chronological message chains with author attribution, timestamps, and channel metadata. We flatten the message history into individual Freshdesk conversation entries preserving the original timestamp, author (agent or customer), and message body. Public agent messages become Freshdesk replies; internal agent notes map to Freshdesk notes on the ticket. The ticket is created in Freshdesk before any conversation entries are written so that the parent_id reference resolves correctly.

Thread Tickets may carry PSA-linked fields when the ConnectWise integration is active. We map Thread Ticket status, priority, assignee, and PSA reference fields to their Freshdesk equivalents. The Freshdesk Ticket status field maps from Thread status, and Freshdesk Ticket priority maps from Thread priority. Assignee mapping resolves Thread agent IDs to Freshdesk agent IDs via the User mapping built during the user phase.

Thread User records (display name, email, role, and team assignment) map to Freshdesk Agent records. We map role assignments so that agent permissions carry over post-migration. If a Thread user has no corresponding Freshdesk agent account, we hold the record in a reconciliation queue and flag it for the customer to provision before the record import resumes.

Thread Teams group agents and set routing rules. We create a crosswalk table during the mapping phase that maps Thread team names to Freshdesk Group names. Freshdesk Groups serve as the routing target for automations, so the group crosswalk must be finalized before any automation rebuild documentation is written.

Thread Response Templates are short-form canned replies with shortcut codes. We export template content and shortcut codes and map them to Freshdesk Canned Responses. Freshdesk supports canned responses at all paid tiers, but the import requires the Freshdesk API or manual CSV upload with UTF-8 encoding. We flag any templates with character counts that exceed Freshdesk's 600-character default limit.

Tags applied to Thread tickets and reviews allow categorization and reporting. We export the full tag vocabulary and map it to Freshdesk Tags, which are applied at the ticket level. Tags that exceed Freshdesk's 30-character limit are truncated with a note in the migration report, and the original Thread tag value is preserved in a custom field for reference.

File attachments within Thread conversations require separate extraction and re-hosting because Thread stores files on its CDN. We download all attachments, rename them with conversation identifiers, and upload them to Freshdesk's attachment endpoint during the conversation import phase. Freshdesk's attachment API accepts files up to 15 MB per file and 25 MB per ticket; we flag any attachment that exceeds these limits for manual handling.

Thread allows custom fields on tickets and conversations. These vary by account configuration. We export the complete custom property schema alongside the data, then create matching Freshdesk custom fields via the API before data migration begins. Freshdesk requires custom fields to be declared as such in the API request body using the cf_ prefix convention; we apply this during the schema design phase so that data load does not fail due to unrecognized field names.