Skip to main content
All CollectionsFor WebQA & Debugging
Troubleshooting issues & the killswitch
Troubleshooting issues & the killswitch
Updated over 11 months ago

Seeing issues on your site and need to know if it's because of an experiment? This article is for you.

Here, we will detail steps to be done in a precise order, that will guide you through working out if the issue is coming from the Optimize platform as a whole or not, and down to which experiment has the issue.

Is the issue severe?

Firstly, an important question is if the issue is severe or not.

Every company has their own line for what is severe, but at the most extreme end, the issue is impacting your ability to trade. Key pages not loading, broken add-to-bag buttons, broken checkout, etc.

If so, please follow these steps:

Checking if the issue comes from Webtrends Optimize

The easiest way to do this is simply to block all network requests from Webtrends Optimize.

If you are able to see the issue on a browser like Chrome, you can open devtools (right-click on the page and click inspect), and then:

  1. Head over to the network tab

  2. Filter requests by "webtrends". Refresh the page at this point.

  3. Right-click on the wt.js file

  4. Click "block request domain"

  5. Hard-refresh the page - Ctrl/Cmd + Shift + R

  6. Observe that wt.js has been blocked (it will be red) and see if the issue persists.

If it does persist, the issue is almost certainly not due to the platform.

You can make absolutely sure by clearing cookies and localstorage, and then browsing through the website.

If you still see the issue, this is conclusive proof that the issue doesn't derive from Webtrends Optimize, which you can relay back to your IT team so they can continue exploring other avenues.

Use the local killswitch

An alternative approach is to add the following to your URL, which will stop any experiments from running:

?_wt.context=optimize:killswitch

If this solves your problem, one of your experiments is likely causing the issue.

Identifying the bad experiment

If you don't have it already, installing the Force Experiment bookmark is a good start.

With the Force Experiment installed, you can remove yourself from particular variations and see if the problem goes away. Put yourself in the control group for each test you belong to one by one, if/until the problem goes away.

This trial-and-error method will help narrow down exactly which test is causing you issues.

Did this answer your question?