Every Test Should Have a Goal
Goals are metric-based objectives that you are aiming to maximize or minimize with an A/B Test. When you add a goal, Leanplum will estimate how long the test will take to reach a statistically significant result.
Adding a goal:
In the A/B test environment, click the Add button to set a new goal, this will open the goal selector window. You can select any of the out-of-the-box Leanplum metrics or choose a goal based on one of the custom events or states that you are sending to Leanplum.
How goals work:
In order to calculate significance, you’ll need to specify whether you are hoping to maximize or minimize each goal metric added. All goals must be either a percent or a ratio. Nominals are not supported as goals and will be greyed out in the goal selector window.
Reaching your goals:
Add as many goals as you like. However, we recommend adding only metrics that will truly make or break this test. If your audience is sizable, a test with multiple goals will return statistically significant results faster than a smaller target audience, so be mindful of how long you wish to run your test for.
Analyzing your goals:
For metrics that you will want to see for all A/B tests (and messages), you can save time by adding them to your Favorites section in the analytics center. Favorites are great for any secondary metrics that may not necessarily be essential to the success of the test but are still important to take into consideration. For example, if you were testing a new color on the order checkout button, your goal might be to maximize purchases. Some favorited metrics you might want to keep your eye on are 3-day retention, percent of daily users who purchased, and time spent in app.
See our blog for a comprehensive A/B testing checklist.