Start and stop the A/B test
In this final tutorial step, start the test manually.
Website users are then served the page variants, and factual data about their interactions is collected.
For the test to work, leave the test running for long enough to generate traffic and serve the different variants.
Artificially generating traffic is not achievable in the scope of a Tutorial, so the following describes the workflow used when a test is run.
Before you start
Complete the other configuration steps up to this point in the tutorial.
Read the documentation about viewing test results in the Optimization documentation.
Steps
The following procedure assumes the Conference CTA Squiz Optimization tab — the active browser tab at the end of the previous tutorial step — is still the active browser tab.
To run the A/B test:
-
Click the Start A/B test button.
-
Click Start in the Start the A/B test? dialog that presents.
-
The test Summary section shows the Running status and information about when the test started, how long it has been running, and how many views have been recorded.
In the Content Management tab, the presentation of the Conferences content page does not obviously change.
If the A/B testing menu in the Actions toolbar is selected, however, the status presents as RUNNING.
-
Leave the test running for long enough that visitors can view the test and interact.
A running test can be paused if necessary.
Pausing a test can, however, invalidate the test findings.
-
Once satisfied with the test duration, switch back to the Conference CTA Squiz Optimization tab, if necessary.
-
Click menu:…[Finish] to manually end the test.
-
Click Finish in the Finish the A/B test? dialog.
-
The test Summary section shows the Finished status.
-
Wait for the results to collate.
-
Analyze the results of the test after the results collate.
-
Disconnect the test from the Conferences page and select the most successful variant from the results.
If you installed Javascript to facilitate a click-based A/B test, remember to remove this as well.