How to Split Test Collection Page Redesigns?
If you have a content management system (CMS) with a collection page like site.com/collection-a and you want to test a redesigned version of that collection, you have a couple options.
One method is to duplicate the entire collection, creating a second URL like site.com/collection-b. Then edit the design of the duplicate collection. Now you need to split traffic between the two URLs to test the new design.
This is where a tool like Google Optimize used to come in handy, allowing you to easily split traffic between two pages. However, since Google Optimize has sunsetted, you need another solution.
How to create a new test
A simple way to split traffic between site.com/collection-a and site.com/collection-b is by using redirection rules with wildcards in Mida.so. Here's how:
- Create a 50/50 split redirection rule with wildcards:
- Control: site.com/collection-a/*
- Variant 1: site.com/collection-b/*
- This rule will automatically redirect 50% of traffic on any page in collection-a to the equivalent page in collection-b.
For example, site.com/collection-a/page1 would split traffic between itself and site.com/collection-b/page1.
Using wildcards in the redirection rules maps all the dynamically generated paths to the test collection automatically. This makes it easy to run a controlled experiment measuring the performance of your collection page redesign.
Why You Should Split Test Collection Redesigns
Split testing a new collection page design minimizes risk and guesswork. Without testing, you may assume your redesigned page will increase metrics like conversion rate or average order value. But launching an untested redesign impacts all your traffic at once. Sales could drop unexpectedly or users may find the new site harder to navigate.
Split testing allows you to first measure if that new flashy, modern layout actually provides a better user experience and leads to real business results. You can catch UX issues early with only a portion of traffic affected. And by comparing analytics between the two versions, you make data-driven decisions about which design better achieves goals like:
- Increasing conversion rate
- Higher average order values
- More clicks on high-margin products
- Lower bounce rates
The bottom line is split testing reduces assumptions. You can answer questions like "Will this new design actually increase revenue?" without taking big risks. And that means maximizing the performance of your critical collection pages.