The A/B Test History report provides an overview of your A/B test program (experiences with A/B test allocation, and at least 2 variations or 1 variation and a control group). The report helps you answer questions such as:
- What were the most successful A/B tests?
- How many tests resulted in increased conversions?
- How many tests have you run on your site?
- Are there as many tests now as previous quarters?
Downloading the report
You can download A/B Test History report by clicking A/B Test History in the A/B Tests dashboard tile.
Report metrics
The report is a CSV file, in which each row represents a variation of a Test Version within an Experience that uses the A/B test allocation method and has at least two active variations.
| Column | Description |
|---|---|
| Campaign Name | The name of the campaign the experience belongs to. |
| Experience Name | The name of the experience. |
| Version Id | The internal ID of the test version. |
| Start Date | The date the test version started. |
| End Date | The date the test version ended, or Still Running. |
| Days Running | The duration of the test version in days. |
|
Variation Name |
The name of the variation. |
| Impressions | The sum of impressions of the variation in this version. |
| Clicks | The sum of clicks of the variation in this version. |
| Primary Metric | The primary metric is used to determine winning variation for the A/B test. The primary metric affects the Probability to Be Best, Uplift, and Winner Declaration calculations. |
| Normalization Type | The type of the denominator used to normalize the metric. For example, users, session, impressions and purchases. |
| Normalization | The count of the denominator. |
| Metric Totals | The count or sum of the metric. For example, purchases or revenue. |
| Metric Totals Excluding Outliers | The count or sum of the metric, excluding outliers. |
| Probability to Be Best |
The probability of the variation outperforming all other variations. |
| Declaration | Whether or not the variation was declared a winner. This value is available only for test versions that were active after July 1, 2023. |
| Probability to Be Best Excluding Outliers | The probability of the variation outperforming all other variations based on data excluding outliers. |
| Declaration Excluding Outliers | Whether or not the variation was declared a winner based on data excluding outliers. |
| Report | A direct link to the report of the experience. |
Note: Only test versions that were active after October 24, 2020 are included.