A theme experiment compares two Shopify themes against each other on live traffic — your current live theme as the control, and a challenger theme you want to prove. Visitors are split between them, and Cooee reports which one performs better on the metrics that matter to your store.
It is the broadest experiment Cooee offers. The challenger replaces the entire storefront for the visitors who see it, which makes it the right tool for a redesign and the wrong tool for a single page.
How it works
Shopify can hold several themes in your library at once and serve any of them on request. Cooee builds the experiment on top of that:
Control — visitors see your live theme, exactly as they would normally.
Variant — visitors are served the challenger theme instead.
The decision happens in the visitor's browser the moment they arrive, with no server round-trip and no waiting. Your storefront URL does not change, so visitors and search engines see your normal addresses throughout.
Assignment is sticky: once a visitor is placed in an arm, they keep seeing that same theme on later visits. Without that, someone bouncing between two designs would produce meaningless data. Automated traffic such as crawlers and bots is excluded and never counted.
Choosing the right kind of experiment
You want to test | Use |
A whole new storefront design, or global elements such as header, footer, colours, or fonts | A theme experiment |
One page type — a new product or collection layout, for example | A template experiment |
A genuinely different page at its own address | A split URL test |
A theme test answers "is this new design better overall?" It does not tell you which of the twenty differences caused the result. If you need that, test the change on its own with a template experiment.
The typical flow
Build and review your challenger theme in Shopify. Cooee never edits your themes.
Create a theme experiment in Cooee, pick the challenger, and set the traffic split.
Publish, and let it run for at least two full business cycles.
Review the results, decide, and then publish or delete the challenger theme in Shopify.
Do not skip step 4. When Shopify serves a visitor the challenger theme, it remembers that choice in their browser, and stopping the experiment in Cooee does not clear it. If the challenger is left sitting in your theme library, those visitors keep seeing it on return visits. Publishing it or deleting it resolves this for everyone.
Read next
Theme experiment use cases — what is worth testing, and when to pick something narrower.
Setting up a theme experiment — prerequisites and the full walkthrough.
Reading theme experiment results — what each metric means and how long to wait.
Stopping a theme experiment and calling a winner — concluding the test and cleaning up.
Troubleshooting theme experiments — answers to the questions that come up most.