How to run an A/B test on your Shop Now bonus (Returns V2)

How to run an A/B test on your Shop Now bonus amount, comparing a test bonus against your current bonus to measure the impact on retained revenue.

2 min readUpdated Oct 1, 2026

This article explains how to run an A/B test on your Shop Now bonus amount, comparing it against your current bonus to measure the impact on retained revenue.

๐Ÿงช A/B testing for the Shop Now bonus is in beta and not available to all merchants. Contact Swap support to check availability for your store.


Prerequisites

  • Shop Now bonus must be enabled. It can be set to 0%, but the setting itself must be turned on โ€” if Shop Now bonus is disabled, the A/B testing toggle is unavailable, with a tooltip explaining why.

How it works

  • Returning customers are split 50/50 into two variants: Current โ€” your existing configured Shop Now bonus, shown as-is and not adjustable here โ€” and Test โ€” an alternative bonus percentage you define for the trial.
  • Assignment happens per session and does not change for the rest of that session.
  • While a trial is running, the bonus amount cannot be changed and advanced bonus rules do not apply.

Setting up a trial

  1. Go to Returns โ†’ Settings, then open the A/B testing tab.
  2. In the Shop now bonus section, enter the bonus percentage for the Test variant. This field is required.
  3. Click Start trial.
    • The dashboard shows the trial's start date and the date it will end (60 days after starting), along with which day of the trial you're currently on.
    • The button changes to Trial running.
  4. To end a trial early, click Cancel trial and confirm. This stops the trial and adjusts its end date; the change is also recorded in the settings log.

Trials end automatically after 60 days if not cancelled sooner.


How the bonus appears on an RMA

  • A standard Shop Now bonus is shown as usual in the RMA tooltip.
  • A bonus applied through an A/B test is labelled: A/B testing bonus provided: {currency}X(Y%).
  • The bonus amount applied to an RMA is locked in at the time of the return and does not change afterwards, even if the trial's settings are changed later.

Limitations

  • Trial results are a comparison between the two variant groups, not a statistically validated experiment. Treat results as directional rather than conclusive.
  • Only a 50/50 split is currently supported.
  • Result and KPI reporting is not yet available in the dashboard. Trial data is captured for internal analysis, but merchants cannot currently view results in Swap.