Key takeaways
- Read an importer’s “not imported” list first. If nothing on it matters to you, the importer may be all you need.
- Zendesk has no built-in ticket importer. Historic tickets go in through its Ticket Import API.
- Whatever you choose, insist on a sample of your own records, a catch-up run at cut-over and a count on both sides afterwards.
What the vendors’ own importers leave out
| Importer | What its documentation says it leaves out or limits |
|---|---|
| Intercom: Zendesk importer | Aimed at moves under 150,000 tickets; all tickets or none; no side conversations, tags, call recordings, multiselect or lookup fields |
| Freshdesk: Zendesk import | No deleted or spam tickets, macros, triggers, inline images or article comments; attachments above 15 MB skipped; ticket IDs change |
| Gorgias: Zendesk import | Up to two years of the most recent data; users with no ticket activity are not imported |
| Help Scout: import tool | Knowledge base, saved replies, rules and custom fields are not imported |
Moving into Zendesk always involves its API: the data importer takes users and organizations from CSV, and tickets go in through the Ticket Import API, which keeps original dates and does not run triggers on imported tickets unless they are closed.
A decision table
| Your situation | Native importer | Self-serve tool | Managed service |
|---|---|---|---|
| Volume inside the importer’s documented range, simple fields | Good fit | Possible | Usually more than you need |
| Only some history: a date range, one brand, open tickets | Often not possible | Usually possible | Good fit |
| Team keeps working during the move and needs a catch-up | Re-run, not a catch-up, on most importers | Check the tool offers a delta | Good fit |
| Need to undo a run if the mapping is wrong | Rarely offered | Check before you buy | Ask for rollback per run |
| No time on your team to own the mapping and testing | Good fit if the data fits | Needs an owner on your side | Good fit |
Where DataGrine fits
DataGrine offers both of the last two routes: a self-serve platform your own team runs, or a migration handled end to end by our migration team, with 24/7 support either way. Both come with the same controls:
- Read-only access to the source, so the old helpdesk is never changed.
- A written field mapping, including every status, dropdown value, agent and group.
- A sample run on your own records, signed off before the full run.
- A delta sync at cut-over and rollback per run.
- An itemised reconciliation report, with every record that did not travel and why.
- Encrypted credentials and a DPA available. More on how we handle data.
If you want a second opinion on whether your data fits an importer, send us your source and destination. If it fits, we will say so.