Importing customers from another helpdesk

You probably aren't starting from zero. If you're switching from Zendesk, Intercom, Help Scout, or Pylon, your customers and ticket history are already there. Ochre's import flow brings both over at once.

What gets imported

For each supported source, Ochre pulls:

• Customers: email, name, profile picture, and source-side IDs (preserved as aliases).
• Tickets: every conversation, including all messages, internal notes where accessible, status, assignee, tags, and timestamps.
• Attachments: stored in your workspace and re-linked.
• Original ticket numbers: preserved as conversation aliases so a customer who emails about "ticket #48201" still threads correctly.

What doesn't import: macros, triggers, automations, and channel routing config. Those are intentional. They rarely map cleanly across helpdesks, and copying them would carry over half-broken behavior. You'll set up routing-rules and tags afresh.

Submitting an import

The import flow lives under workspace settings as Helpdesk Import Requests. Pick a source, then provide the credentials it asks for.

Credentials

Each source needs different things:

• Zendesk: subdomain, admin email, and an API token.
• Intercom: a Personal Access Token from a workspace owner.
• Help Scout: an OAuth connection (no key to paste).
• Pylon: an admin API key.

Credentials are encrypted at rest and accessible only to the import worker. They are not visible to any user once submitted, including admins. After the import completes, you can have Ochre rotate or delete them.

Source-specific notes

For source specifics, see migrate-from-zendesk, migrate-from-intercom, migrate-from-helpscout, and migrate-from-pylon.

How the worker runs

After you submit, an import worker picks up the request and starts pulling pages from the source's API. It works through customers first, then tickets, then attachments. Each phase reports progress.

Throughput depends on the source's rate limits, not Ochre's. A 100k-ticket Zendesk usually finishes overnight. Intercom is slower per page, so big workspaces can run longer. The status page shows estimated completion based on observed throughput.

You can keep using your old helpdesk during the import. The worker is read-only on the source side, so your team's day-to-day isn't disrupted.

What happens to existing customers in Ochre

If you've already received conversations in Ochre before importing, the importer will merge by email. If a customer in your old helpdesk shares an email with one in Ochre, their tickets and aliases attach to the existing record. New customers are created from scratch.

For ambiguous matches (different name, same email, different profile picture), Ochre keeps both as aliases on the same customer rather than guessing. You can clean up after the fact with customer-merge.

Custom fields on import

Most helpdesks have their own custom fields. The import wizard lets you map source fields to existing or new Ochre custom fields. Mapped values come over with the customer record. Unmapped values are dropped, with a CSV export available if you want to preserve them externally. See custom-fields.

Verifying after import

Once finished, the import status page shows totals: customers imported, conversations imported, attachments imported, and any items skipped with reasons.

Spot check:

1. Search for a known customer and confirm their conversations are present.
2. Open a couple of old conversations and confirm message order, attachments, and assignee.
3. Send yourself an email referencing an old ticket number; confirm it threads to the imported conversation via conversation-aliases.

Re-running and incremental imports

Imports are idempotent. Running the same import twice won't duplicate customers or tickets; it'll backfill anything new since the last run. Useful when you want to do a first import for testing, then a final import on cutover day.

When something goes wrong

If a source's API is rate-limited or down, the worker pauses and resumes automatically. If credentials expire mid-run, you'll see an error on the status page and a button to re-supply them. For stuck imports, see migration-stuck.

After cutover

When you're confident the import is clean, point your inbound email to Ochre and turn off the old helpdesk. From that point, every new conversation lives in Ochre, but the history is right there alongside it.