Because the whole point is to need fewer of them. A returns team that stops re-keying orders shouldn’t pay more for the privilege.
One returns team getting started
A full returns operation
Multiple brands, regions or clients
| Feature | Starter | Growth | Enterprise |
|---|---|---|---|
| Return requests | Up to [1,000] / mo | Up to [20,000] / mo | Unlimited |
| Workspaces | [1] | [3] | Unlimited |
| Intake channels | [1] | All | All |
| Order-system connection | Import file | Connector, API or file | Connector, API or file |
| Custom field mapping | — | Included | Included |
| Dedupe, linking & audit | Included | Included | Included |
| Customer questionnaire | Included | Included | Included |
| Status sync & ticket updates | Included | Included | Included |
| Sandbox workspace (all writes off) | Included | Included | Included |
| SSO / SCIM | — | — | Included |
| Regional data residency | — | — | Included |
| Support | Shared channel | Named contact + SLA |
Limits in brackets are placeholders until pricing is final.
One inbound return request that becomes a return order — an RMA — in your system. A follow-up linked to an existing return doesn’t count twice — and neither does a duplicate Returnberry caught.
Included, and metered per workspace so you can see it. Structured requests are read without AI at all, which is why the bill stays boring.
Yes. Run a workspace with every outbound write forced off: it reads, parses, dedupes and prepares orders, and contacts nobody.
We’ll run them through and show you the orders that come out.