A template experiment lets you A/B test two versions of a Shopify page template — your existing template versus an alternative you design — by splitting live traffic between them. It's the most surgical way to test on Cooee: instead of swapping your whole theme, you test one page type (product, collection, homepage, and so on) in place.

Shopify lets a single theme hold more than one template for the same page type. The extra ones are called alternate templates and are identified by a suffix — for example product.daypack.json is the daypack alternate of your product template. Shopify renders an alternate template whenever the URL carries a ?view=<suffix> parameter.

Cooee builds the experiment on top of this native mechanism:

Assignment happens in the visitor's browser the moment they land on a targeted page. It is deterministic (the same visitor always lands in the same arm) and needs no server round-trip, so there is no added latency and no flicker. Because assignment relies only on the URL parameter, there is no persistent cookie to clean up — when you stop the experiment, Cooee simply stops adding ?view= and everyone sees the control again.

Template experiments are available for every standard Shopify page type:

A template powers many pages at once — one product template typically serves every product. So a change you make to a product variant template is tested across all product pages in scope, not one product at a time. You can narrow the scope to specific URLs (see Setting up a template experiment).

A template experiment only controls the content between your header and footer. It cannot change:

To test global elements or a completely different look, use a Theme A/B test instead.

You build the alternate template yourself in Shopify — Cooee never edits your theme. The typical flow is:

See Setting up a template experiment for the full walkthrough and Managing your variant templates for creating and naming templates in Shopify.