Usage:AB Tests

From Fuseboxx Wiki
Jump to: navigation, search


A/B Testing utilizes the FuseBoxx™ App Parameter system to create custom experience tests for your app's users. Data from these users can be compared (the users are separated by which test they fall in), and when the tests are complete, FuseBoxx™ can push the Parameters to the Standard experience if you so choose. We can set up a number of variants to be tested, and we can also set up how many users are put into each variant.

NoteBubble.png

For more information on App Parameters, if you are developing on iOS Click Here, or if you are developing on Android Click Here. You can also view our A/B testing guide Here

Creating a Test

To begin A/B Testing open the A/B Testing page (Toolboxx Tab > A/B Testing), here you will see a list of all your tests (if they exist), to create a test press the Test Details subsection, enter the basic information of your test.

In Participant Inflow Controls you will want to Allow Manual Inflow Controls, and enter a User Cap (The max # of people in allowed in the tests) and End Date (when the Test is over and users go back to the regular app experience).

Next you need to decide whether you are going to put New Users into the test or Users from an existing Cohort, this will depend on what you are testing for, you may only want to conduct a test on users who have already paid into your app.

Comparing Test Results

When the test is complete you can compare the results of the test in Test Results. Here you can select which test to view and what the Y-axis you are comparing the users against from the dropdown menu's. You can also compare the tests at a glance in the All Tests section.

ABTest Results.png

Here we are comparing the revenue earned by the three variants of the test.

You can also compare the results from an A/B Test By Time, here you will have to select which test you want to view and the metric to be displayed.

ABTest By Time.png

Here we can see that players in Test variant C contributed more to the revenue of the app, and we can see where the trend started.