Import historic support tickets

Historic support tickets can be valuable: they explain a customer's previous problem, show what was promised and reveal repeated product faults. They can also make a new help desk look broken if thousands of old messages appear as open cases, duplicate contacts or misleading performance data.

The question is not “Can we import everything?” It is “Which history must be available to answer customers safely, and what is the least risky way to make it available?” A selective import plus a searchable archive is often better than a technically complete but confusing copy.

Decide the import scope first

Ask the organisation's data owner what should be retained and who should access it. A migration is a poor reason to extend a retention period by accident. If old messages are not required in the live queue, document where the archive lives and how an agent can search it.

Inventory source systems and identifiers

List every source: shared mailbox, prior help desk, CRM, web form and any support conversations in a chat tool. Note whether the export includes message bodies, thread ordering, attachments, customer identifiers, private notes, tags and original status. An export that contains one row per ticket but not the conversation is not a complete ticket history.

Preserve a source identifier on every record. It lets the team reconcile duplicates and return to the original if a destination field looks wrong. If a customer appears under several email addresses, decide whether to merge identities or keep them separate pending review. Automatic merging by name alone is risky.

Count records by status and by a few time periods before changing them. Record the number of attachments and the size of a sample. These baseline totals are essential when the import tool reports “success” but silently skips some fields.

Map fields in a small sample

Choose around twenty varied cases: one open, one waiting, one reopened, one long thread, a customer with several tickets, a ticket with a private note, and several attachment formats. Map each source field to a destination field. Mark any transformation, such as source “pending supplier” becoming destination “waiting”, and write down what information is lost.

Check the date and time zone explicitly. A timestamp shifted by a day can change whether a promised response appears overdue. Do not rely only on a pretty date in the interface; inspect the raw export or import report where possible.

Test an old customer reply. If a person responds to a thread from before migration, will it attach to the imported ticket, create a fresh ticket or land in the old mailbox? Any of these can be workable if agents know the rule. Silent loss is not.

Understand the provider's import route

Zoho Desk's current import guide describes importing module records, including tickets, with CSV files and mapping. Edition and permission requirements still need checking on the account being used. Do not confuse this with an older help article showing a different interface.

HelpDesk's migration documentation describes mapping fields and using a small test import before a larger move. Its route may involve provider assistance or API work depending on the source. Ask what the service will and will not bring across, and whether there are charges or capacity limits.

Zendesk's migration guidance directs ticket import to the API. It also says imported tickets do not receive historical first-reply or first-resolution metrics and that triggers do not run on imported tickets. This is a strong example of why a ticket record and its historical reporting are separate things.

HubSpot's import tool overview explains record and association mapping. If moving conversation history into a CRM-led service workspace, test whether ticket properties, messages and linked contacts all arrive in the form agents need; a ticket-object import alone may not recreate a service conversation. For Freshdesk, get the precise route confirmed for your source and plan rather than inferring it from its contact import screens.

Protect customers from import side effects

Before a test, review acknowledgement emails, notifications, assignment rules, SLA timers and any connected apps. An old ticket arriving as a new record might otherwise be treated as a fresh customer request. Even where a vendor suppresses triggers on import, integrations outside the help desk may react to the new object.

Use a non-production destination or a very small controlled sample when available. If production testing is unavoidable, choose records owned by internal staff and confirm the mail settings first. Keep a record of the exact import batch so it can be inspected, corrected or rolled back according to the provider's capabilities.

Do not let imported closed tickets flood the normal queue. Tag or segment them clearly. Decide whether agents may edit old history, and whether customer-visible replies on imported records are allowed during the cutover.

Reconcile beyond the success message

Compare the destination with the source totals, then inspect the representative cases manually. Can an agent see every message in order, identify the original requester, open attachments and distinguish private from public content? Check a case with several people copied on the thread. Check a long case whose old status changed repeatedly.

Make a discrepancy register: missing records, merged customers, unmapped status, shifted time, broken attachments and unexpected notifications. Resolve material discrepancies before a larger import. If the import tool reports skipped rows, keep the report and investigate the pattern; do not merely rerun the same file and create duplicates.

After a full import, reconcile totals again by status and sample the oldest and newest records. Look at live reports: did historical records suddenly increase the current backlog, affect response averages or trigger overdue alerts? Filter them out if they are reference history.

Keep old and new reporting separate

A help desk often measures first response from its own ticket events. An imported conversation may display the original message but lack the original metric events. A current report can therefore show no first response, a response from migration day, or a different denominator from the previous system.

Mark a reporting boundary at cutover. Report pre-migration history from the old system or a preserved export; report new work from the new system. Only combine periods when definitions, time zones and event data have been verified. The small-team reporting guide sets out practical definitions.

When not to import

Do not import merely to avoid telling agents where the archive is. If the destination cannot preserve a conversation's meaning, maintaining read-only access to the old source can be safer. Likewise, a few active cases can be recreated manually with a concise internal summary and a link to the archive, provided the customer-facing promises remain intact.

Importing everything is especially hard to justify if the material is rarely referenced, contains sensitive attachments, lacks clean identifiers or requires a paid migration project out of proportion to the benefit. Document the decision so a future agent knows where historic answers live.

Product fit and next step

Choose Zoho Desk when its documented import route and target edition suit the source. Consider HelpDesk when a focused queue and supported migration route are more valuable than broad customisation. Zendesk can handle an API-led project when the organisation can plan the mapping and accept metric limitations. HubSpot deserves scrutiny when associations with the customer record matter. Freshdesk may fit the wider service workflow, but the exact ticket-import method is a vendor question to settle before purchase.

The import should be one workstream inside the full inbox migration, never the only measure of success. The real outcome is whether agents can answer current customers accurately on launch day.

Sources and further reading

• Zoho Desk: importing data.
• HelpDesk: migrate from another ticketing system.
• Zendesk: migration options and metric limits.
• HubSpot: understand the import tool.

The selection framework and product-fit conclusions are Support Sorted's editorial guidance.