What you'll build

Two checkout setups — one that leads with "Free delivery over €75" messaging and one that leads with estimated delivery days — with an experiment that splits incoming traffic 50/50 between them. You will be able to observe which variant drives higher checkout conversion in your analytics platform.

Before you start

• You have two distinct checkout setups already created and configured with their respective shipping options and messaging. If not, create the second setup by cloning your existing one and adjusting the option names and descriptions.
• Your analytics platform (Google Analytics, Segment, Mixpanel, etc.) is connected to your checkout and can receive custom properties or session attributes. Shipit does not provide a built-in results dashboard.
• You understand that experiments affect the entire checkout setup, not individual shipping options.

Step 1 — Prepare the two checkout setups

Variant A — Free Shipping Emphasis:

Configure the shipping option names and descriptions to lead with the free-shipping threshold. For example:

Variant B — Speed Emphasis:

Configure the same options with delivery time as the primary message:

Both setups should use identical pricing and carrier services — you are testing messaging, not logistics.

Step 2 — Add an experiment to Variant A

Open Variant A (the free-shipping setup) in Delivery Checkout → Checkout Setups. Navigate to the Experiment section and click Add Experiment.

Leave the experiment inactive for now. You will activate it after validating Variant B in the next step.

Step 3 — Validate Variant B before going live

Before running a 50/50 split, route a small percentage of traffic to Variant B to confirm it is not broken.

Tip: Starting at 90/10 lets you catch configuration errors (missing shipping options, pricing mistakes, broken translations) with minimal customer impact. Monitor for 24–48 hours, then move to 50/50 once you are confident Variant B is working correctly.

Step 4 — Move to a 50/50 split

Once Variant B is validated, update the experiment:

Save the experiment. Shipit will now assign new checkout sessions to buckets and each customer will see the corresponding setup.

How it works

When a customer's checkout session starts, Shipit automatically and consistently assigns each customer to either Bucket A or Bucket B — the same customer always sees the same variant across sessions. A customer assigned to Bucket A sees the Variant A setup; Bucket B sees Variant B.

Shipit passes which variant a customer sees as part of the context sent to your checkout provider. Your analytics platform can read this value and use it as a dimension when comparing conversion rates between the two groups.

Warning: Shipit does not analyse experiment results. You must set up conversion tracking in your own analytics platform and segment results by bucket label. Without this tracking in place before you start, you will have no data to act on.

Step 5 — Ending the experiment

Once you have enough orders to draw a reliable conclusion about which variant performs better, update the experiment:

1. Set Is active to false.
2. If Variant B won: make Variant B your primary setup and update the experiment's Bucket B reference accordingly, or simply delete the experiment and use Variant B directly.
3. If Variant A won: delete the experiment and continue using Variant A.

Warning: Disabling the experiment does not automatically redirect all traffic to one variant. Confirm which setup is set as the default for your checkout configuration after ending the experiment.

Testing it

1. In two separate browser profiles (to simulate different sessions), open your checkout and add the same item to each cart.
2. Verify that one session shows free-shipping-first messaging and the other shows speed-first messaging.
3. Confirm your analytics platform receives the bucket label as a session property.
4. Check that pricing and carrier availability are identical across both variants — only the wording should differ.