> For the complete documentation index, see [llms.txt](https://docs.elevateab.com/elevate-helpcenter/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.elevateab.com/elevate-helpcenter/product-testing/product-testing-overview.md).

# Product Testing Overview

What product experiments do in Elevate A/B Testing, how product duplication works, and what you need to prepare before you start.

***

A product experiment compares one product listing against another. Different title, different description, different bundle, different images, or an entirely different offer.

It works by pairing each product you are testing with an alternative product, then showing each visitor one or the other.

### How it works

1. You select the products to test, up to 500.
2. Each selected product becomes a **group**. Every group needs a product to test against.
3. For each group you either pick an existing product or have Elevate A/B Testing duplicate the original so you can edit the copy.
4. Visitors are split, and each one sees the product assigned to their variation.
5. Elevate A/B Testing tracks behaviour and attributes conversions to the product each visitor saw.

Duplicated products are created **Unlisted**, so they do not appear in your storefront navigation, search, or collections. Visitors reach them only through the experiment.

### What you can test

| Change            | Question it answers                                                                                                                                              |
| ----------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| Product title     | Does a clearer or more descriptive name sell better                                                                                                              |
| Description       | Does a different length, structure, or angle convert                                                                                                             |
| Bundle or offer   | Does a bundle beat the single item                                                                                                                               |
| Images and media  | Handled better by a [Product Image Testing Overview](https://claude.ai/elevate-helpcenter/test/product-image/overview) unless the images come with other changes |
| The whole listing | Does a rebuilt product page beat the current one                                                                                                                 |

### When to use it

| Use a product experiment when                                 | Use something else when                                                                                                                           |
| ------------------------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------- |
| You want to change what the product is or how it is described | You only want to swap photos. Use a Product Image Experiment.                                                                                     |
| You are testing a bundle against a single item                | You want to change the price. Use a Price Experiment.                                                                                             |
| The change needs a genuinely different product record         | You want to change the layout of the page. Use a [Page Testing Overview](https://claude.ai/elevate-helpcenter/test/page/overview).                |
| You want to test copy across many products at once            | You want to edit text without duplicating anything. Use a [Content Testing Overview](https://claude.ai/elevate-helpcenter/test/content/overview). |

### Requirements

* Elevate A/B Testing installed on your store.
* The Elevate A/B Testing theme extension enabled.

Product testing works on all Shopify plans.

### What duplication costs you

This is the type with the most housekeeping, because a duplicated product is a real product in Shopify and it starts life without everything the original has accumulated.

| Carries over                                | Does not carry over                                |
| ------------------------------------------- | -------------------------------------------------- |
| Title, description, images, price, variants | Product reviews                                    |
|                                             | Inventory and fulfilment settings from apps        |
|                                             | App configurations tied to the original product ID |

Three things need attention before you launch, and Elevate A/B Testing walks you through all three in the review before launch:

**Reviews.** A product with no reviews against one with two hundred is not a fair comparison, and the missing reviews will lose you the experiment on their own. Reviews have to be copied by your review app's support team.

**Inventory and SKUs.** Variations should be recognisable as the same product to your other systems. Mismatched SKUs cause problems in fulfilment, reporting, and any app that reads product identifiers.

**Sales channels.** Duplicated test products should be limited to the online store. Left on other channels, they can appear in your Facebook or Instagram catalogues as separate listings.

None of this is difficult. All of it is easy to forget, and forgetting it produces a result that looks real and is not.

### Next steps

* [Create a Product Experiment](https://claude.ai/elevate-helpcenter/test/product/create)
* [Preview and QA a Product Experiment](https://claude.ai/elevate-helpcenter/test/product/preview-and-qa)