Skip to main content
menu_icon.png

 

x
Optimizely Knowledge Base

Add bounce rate or exit rate metric to an experiment

This article is about Optimizely X. If you're using Optimizely Classic, check this article out instead.
 
relevant products:
  • Optimizely X Web Personalization
  • Optimizely X Web Experimentation

THIS ARTICLE WILL HELP YOU:
  • Add bounce rate and exit rate metrics to your experiments and campaigns, whether they're already running or not

Bounce rate and exit rate metrics are configured on pageview events. You can add them to new experiments or experiments that are already running.

Add bounce rate or exit rate metrics to a new experiment

  1. From the Experiments menu, click Create New... and select A/B TestMultivariate Test, or Personalization Campaign.

  2. Enter an experiment name, choose the pages you’re testing and audiences to include, and add variations for the test.

  3. Under Metrics, select the desired pageview event.

  4. Enter a descriptive name for the metric, such as “Bounce rate for homepage.”

  5. Select Decrease as the winning direction (the goal is typically to decrease bounce and exit rates by making your website more engaging).

  6. Select Bounce Rate or Exit Rate from the dropdown list of measurements.

    bounce-rate-metric-setup.png

  7. Click Add to Experiment (or Add to Campaign).

  8. Click Create Experiment.

Add bounce rate or exit rate metrics to a running experiment

  1. Select the experiment that you want to add the metric to.

  2. Under Manage Experiment, click Metrics.

  3. Select the desired pageview event.

  4. Enter a descriptive name for the metric.

  5. Select Decrease as the winning direction (the goal is typically to decrease bounce and exit rates by making your website more engaging).

  6. Select Bounce Rate or Exit Rate from the dropdown list of measurements.
    bounce-rate-metric-setup-2.png

  7. Click Add to Experiment (or Add to Campaign).

  8. Click Save.

 When you add bounce or exit rate metrics to a running experiment, they will be calculated retroactively, back to the beginning of the experiment.