Split URL testing

  • Updated

Marketers use A/B testing to create personalized experiences that resonate. By methodically testing the effectiveness of messaging, calls to action, and landing pages, marketers can generate real-world data to help them maximize conversions and create delightful user experiences. But this often requires help from developers, who may not always be immediately available to assist.

With Amplitude’s split URL testing feature, you can design, deploy, and analyze A/B tests that involve or require redirecting visitors to another URL, without extensive developer involvement. It’s a fast and easy way to assess the effectiveness of redirects for achieving goals like increasing conversions or improving the overall user experience.

Split URL testing works well if you’re building different versions of your page or site on a CMS like Wordpress. In these cases, your different URLs and their associated pages are the variants that Amplitude Experiment will track.

You can use split URL testing with both standard A/B tests and multi-arm bandits.

To use split URL testing, you’ll have to add a script tag for Amplitude’s Experiment SDK to the site you want to test, in the <head> section. You will also need the Amplitude Analytics SDK present on your site, for event tracking.

Set up split URL testing

To set up a URL redirect test, follow these steps:

  1. In Amplitude, navigate to Create > Experiment > Use a URL Redirect. 
  2. In the Variants panel, add each URL you want to test as a separate variant. 
  3. In the Analysis Settings panel, copy the code snippet to paste between the <head> and </head> tags of your site. Also, make sure your site has the Amplitude Analytics SDK installed, to properly monitor your exposures.
  4. For the rest of the setup process, follow the guidance provided in our experimentation workflow documentation.

Preview and test

As with feature-level experimentation (using flags), you can test the behavior of the redirect before deploying it. The preview is available after you’ve attached URLs to all your variants. 

When you preview and test, you can specify which variant you want to verify for the redirect behavior.

Configuration limits

There are limits imposed on experiment configuration when using visual experimentation and Amplitude’s low-code implementation:

  • Evaluation mode: This is limited to local, in order to optimize test performance and minimize any latency impact to end-users.
  • Bucketing unit: Because evaluation mode is limited to local, this must also be limited to User.
  • Keys: Because evaluation mode is limited to local, this must also be limited to deviceID
  • Audience: This is limited to all users, because the URL redirect logic will trigger before user tracking loads on the site.
  • Deployment: This is limited to the project API key, in order to simplify setup requirements.