A/B Testing Field Reference
| Field Name | Definition |
| A/B Test Name (required) | Name your test. |
| Description (required) | Describe which content will be compared. |
| Hypothesis (required) | Explain what you predict to happen when you run the test, what the outcome will be, and why it will happen. |
| Success Metrics and Results |
Select one or multiple metrics to measure, in support of your hypothesis. Results are reported using these metrics. |
| Page Views | Number of times the visitor viewed the page. |
| Bounces/Bounce Rate | Percentage of visitors who leave ("bounce") rather than continuing to view other pages within the same website. |
| Time on Website/Time Spent (Avg) |
Amount of time the visitor spends navigating your website. Results will show an average of the visitor's time spent. |
| Clicks/Click-Through Rate |
Number of times a unique visitor clicks anywhere on the block, including links. Results will show both the number of clicks and the click-through rate. Click-through rate is the percentage of visitors who click compared to the number of visitors who view the page. |
| Unique Visitors (Results) | A new visitor who sees any A/B test content for the first time. Visitors are assigned several first-party cookies that are used to uniquely identify them. |
| Audience Participation |
How many of your visitors should enter this A/B test? If you select 70%, then 70% of all your visitors will enter this A/B test and see either the original version or any variation. The other 30% will not enter the A/B test and will always see the original version. |
| Start Date Time |
If specified, this A/B test will start automatically on the start date. Leave the field empty if you want to start the A/B test as soon as this test is placed on a page. The A/B test selects the time zone in the following order:
|
| End Date Time |
If specified, this A/B test will end automatically on the end date. Leave the field empty if you want to finish the A/B test manually. The A/B test selects the time zone in the following order:
|
| Advanced Settings |
We use the Confidence Threshold and the Minimum Detectable Effect to calculate the number of visitors you need before you can be confident about the results. Run the A/B test long enough to ensure the detected effect is not due to randomness. We recommend running the test for at least two weeks. |
| Confidence Threshold |
The goal of an A/B test is to make sure you collect enough data to confidently make changes based on the results. The higher the number, the more likely it is that the results are real, repeatable, and not due to random chance. |
| Minimum Detectable Effect |
Minimum detectable effect (MDE) is the relative minimum improvement that you expect to detect. For example, if the conversion rate of a goal is 10%, and you expect a 20% MDE, then a variation will need to have a conversion rate of at least 12% in order to be a winning variation. |