How to Monitor Your Experiment

Monitoring the exposures for all variants and control groups is necessary if you want to monitor the health of your experiment. Diagnostics can be used to validate that your application is recording events as expected for users who are exposed to the experiment.

Hourly Non-cumulative Exposure

You can view the Hourly Non-cumulative Exposure on this page. The effective time range is from experiment creation to 12 hours after running.

Data Diagnosis Tool

You can also check whether the users are assigned to this experiment through the Data Diagnosis Tool.

image (1).png

Type in the user ID and start querying.

image (2).png

We can also generate two user IDs for you to test this experiment, one is assigned to the experiment and the other one is not.

image (3).png

You can copy the user ID and make a query. The diagnosis results will be shown below.

image (4).png

For example:

image (5).png

Detail

Scroll down to the Detail panel, where you will see a stream of exposure below. You can click refresh to see the latest exposure.

image (6).png