Compare the pages people actually use
Split URL testing compares two versions of a webpage hosted at different URLs. Visitors are divided between them, so you can see how each page performs during the same period.
Imagine you run a home renovation business. Your current landing page explains your services and has a “Request a quote” form. You’ve designed a new page with larger photos, shorter copy and a more prominent form. A split URL test shows some visitors the existing page and others the redesign.
To keep the comparison fair, visitors should be assigned randomly and normally see the same version if they return. If one page gets mostly mobile visitors while the other gets desktop visitors, the results might reflect the audience rather than the design itself.
Measure what matters
The new page might look more polished, but that doesn’t mean it will bring in more enquiries. Track the action you want visitors to take, such as completed quote forms or purchases. Compare the conversion rate for each page: the share of visitors who take that action.
Keep the audience and measurement consistent while the test runs. Avoid calling a winner after only a handful of visits; a small difference could be chance. If your setup uses temporary redirects or canonical tags, make sure they’re handled correctly for SEO.
Look beyond the total number of form submissions, too. A page that brings in fewer but better-qualified enquiries may be more valuable than one attracting lots of people who never become customers. When possible, check what happens after someone submits the form.
Decide with evidence
If the redesign earns more meaningful conversions, you have a reason to adopt it. If it doesn’t, look at what may be getting in the way before replacing the original.
Test before making a major website change. A design decision is stronger when it reflects what visitors actually do.