Skip to content

Staging Your Changes

Shadow mode lets you make changes to your shipping rules and test them against real checkout traffic without affecting what customers actually see. Think of it as a staging environment for your shipping configuration.

  1. Enter staging mode

    Click “Enter Staging” in the dashboard. This creates a new changeset — a container for all your staged changes.

  2. Make changes

    Add, modify, or delete rules, methods, and groups. These changes are staged — they’re saved in the changeset but not applied to your live configuration. Customers still see your existing rates.

  3. Watch real traffic

    While you’re in staging mode, every real checkout is evaluated twice: once against your live configuration (which the customer sees) and once against the staged configuration (which only you see). The results are saved as shadow results.

  4. Review the shadow dashboard

    See how your staged changes would have affected real checkouts. Compare live rates vs. staged rates for each checkout. See how many rate differences were detected.

  5. Promote or discard

    When you’re confident the staged changes are correct, click “Promote” to apply them as your new live configuration. Or click “Discard” to throw them away — your live configuration stays unchanged.

A changeset goes through these stages:

StatusMeaning
DraftYou’re making changes. No shadow results yet.
TestingChanges are locked. Real traffic is being evaluated against both live and staged configurations.
PromotedChanges have been applied to the live configuration.
DiscardedChanges were thrown away.

You can only have one active changeset at a time. Promote or discard the current one before starting a new staging session.

When you make changes in staging mode, the following operations are tracked:

  • Create, update, or delete rules
  • Create, update, or delete shipping methods
  • Create, update, or delete product groups
  • Create, update, or delete blended rates

Each individual operation is stored in the changeset. When you promote, they’re all applied to the database in a single transaction.

  1. Make focused changes. Keep changesets small and targeted. A changeset with 2-3 related changes is easier to evaluate than one with 20 unrelated changes.

  2. Wait for enough traffic. The shadow dashboard is most useful when you have a statistically meaningful number of checkouts. For a typical store, 50-100 checkouts gives a good picture.

  3. Check the coverage metric. The dashboard shows how many unique zone/group combinations have been tested. Low coverage means some scenarios haven’t been seen yet.

  4. Use Test Rates too. Shadow mode tests against real traffic, but you can also use the Test Rates panel to simulate specific edge cases that real traffic might not cover.