Test and troubleshoot

Quality assurance is the last step in creating an experiment. This topic covers some approaches to verifying that your variations look and work the way you expect. You should not make changes or attempt to test an experiment after live visitors have been bucketed, as this can affect your results.

Below, we cover the testing tools available for Optimizely Performance Edge experiments.

If you're working in a staging or development environment, start your experiment and use force variation parameters to preview each variation.

You can also push logs to the console by loading the page with ?optimizely_log=info parameter appended to the URL.

In a production environment, you may need to combine multiple methods below to test without exposing your visitors to the experiment.

View the log of changes being applied, as well as logs about experiment decisions, with ?optimizely_log=info.

Many microsnippet logs will print before all tracking client logs.

Preview each variation with a force parameter (e.g. ?optimizely_x=VARIATIONID).

To find the Variation IDfind the Variation ID:

When you're ready to make your experiment live to the world, change your URL targeting conditions to target a page without your test query parameter.

You can also achieve query parameter targeting with a query parameter audience conditionquery parameter audience condition.

When you are ready to make your experiment live to the world, remove the cookie audience condition from the experiment.

You can temporarily disable Optimizely Performance Edge by adding the query parameter optimizely_disable=true to the web page URL. When that happens, an empty JavaScript file will be returned so you can detect if any issues are caused by the Optimizely Performance Edge snippet itself.