Split Tests for Landing Pages
We’ve introduced a new experiment type: Split Tests.
Split tests allow you to compare different landing pages to see which one performs best. Instead of modifying elements on a single page, visitors are randomly redirected to completely different URLs, allowing you to test full-page experiences against each other.
This is particularly useful when testing major design changes, different page structures, or entirely new landing pages.
How Split Tests Work
When a visitor enters a split test:
- The visitor is randomly allocated to a variation.
- Based on the assigned variation, the visitor is redirected to the corresponding landing page URL.
- The experiment tracks performance for each variation to determine which page performs best.
Each variation represents a different page version, hosted at its own URL.
Creating a Split Test
To set up a split test:
- Create the alternative landing pages yourself.
- Host each version on its own URL.
- In the experiment setup, provide the URL for each variation.
Example setup:
| Variation | URL |
|---|---|
| Control | /landing-page-original |
| Variation A | /landing-page-redesign |
| Variation B | /landing-page-video-version |
Visitors will then be randomly redirected to one of these pages when they enter the experiment.
Why Use Split Tests?
Split tests are ideal when you want to evaluate large differences between pages that cannot easily be implemented through visual editor changes.
Common use cases include:
- Testing completely redesigned landing pages
- Comparing different marketing campaign pages
- Experimenting with different layouts or messaging strategies
- Evaluating new page concepts before a full rollout
Because each variation is a separate page, you have complete freedom in how each version is designed and built.
Visual Editor vs Split Tests
Split tests complement traditional on-page experiments created with the Visual Editor.
| Experiment Type | Best For |
|---|---|
| Visual Editor Experiments | Changing elements within an existing page |
| Split Tests | Comparing entirely different landing pages |
This makes it easier to choose the right experiment approach depending on how large the change is.
Test Entire Page Experiences
With split tests, you can move beyond small UI changes and evaluate full landing page concepts. By redirecting visitors to different page versions, you gain insight into which design, layout, and messaging combination drives the best results.
If you’re testing major page redesigns or campaign landing pages, split tests provide a simple and reliable way to measure performance.