Schedule an A/B test

The schedule for an A/B testing determines when and how the test will be started and stopped.

The schedule is configured from the schedule section on the A/B testing configuration screen.

Test start

This defines how and when your A/B test will start. You can schedule a specific date for the A/B test to automatically start, or choose to manually start the test.

Start date

Set the date and time that the A/B test will start from the calendar picker. The start time can be set using the two select dropdowns that appear below the calendar. The A/B test will automatically begin on this date.

The timezone is determined by your web browser and will normally be your local time.
Start manually

When this option is selected the A/B test will only start when you choose the start test option from the A/B testing management screen.

Test duration

This defines when your A/B test will finish. You have a choice to set a specific end date, base the test on a fixed number of visitors, or just let the test run indefinitely until you manually end the test.

End date

The A/B test automatically finishes on this date and time. The end time can be set using the two select dropdowns that appear below the calendar.

The timezone is determined by your web browser and will normally be your local time.
Visitor amount

The A/B test ends when a set number of visitors have been served the test.

Finish manually

The A/B test runs until you manually finish the test.

You can manually finish an A/B test by managing a running test and clicking the …​ button and selecting the Finish option.
Calculate duration

Expanding the calculate duration panel opens a tool to help advise you on an appropriate end date or total visitor amount to use in your A/B test. The calculation is based on some basic site metrics, and you will need to provide your current conversion rate, desired conversion rate and daily visitor numbers.

Variant presentation

If an A/B test is not currently running, Variant A is shown on the page by default.

If an A/B test is scheduled to run in the future, Variant A displays on the page until the test is running.

Similarly, when an A/B test ends, the page returns to displaying Variant A.