# How to Create a Product Test This guide walks you through setting up and launching a single product test using a duplicate product. With this method, you can test entirely different versions of a product against the original — different prices, descriptions, images, bundles, or full alternative offers. > **Just want to test prices?** Use [Price Testing](https://docs.elevateab.com/elevate-helpcenter/price-testing/how-to-create-a-price-experiment) instead — it's much simpler and doesn't require duplicate products. > > **Testing across multiple products?** See [Testing Multiple Products (Grouped)](https://docs.elevateab.com/elevate-helpcenter/product-testing/testing-multiple-products-grouped). *** ### Step 1: Create an Experiment From your Elevate dashboard, go to **Experiments** and click **New Experiment**. Select **Product Experiment**. ### Step 2: Name Your Experiment and Write a Hypothesis Give your experiment a clear, descriptive name that makes it easy to identify later — something like "Premium Bundle Offer Test" or "Holiday Gift Set." We strongly recommend filling in the **description** and **hypothesis** fields. Your hypothesis should state what you expect to happen and why. For example: > *"Bundling our moisturizer with a free travel-size cleanser and raising the price by $5 will increase conversion rate and average order value by giving customers more perceived value."* This documentation becomes valuable when you're reviewing results weeks later and need to remember what you were testing and why. ### Step 3: Select Your Product Choose the product you want to test from your store. * If you've already created a duplicate of this product, select it. * If not, click **Duplicate Product** to create an exact copy, then modify the duplicate however you want — change the price, swap images, rewrite the description, adjust variants, or anything else. The duplicate is your variation. The original is your control. Visitors will be split between them. ### Step 4: Set Traffic Allocation By default, traffic is split evenly between your control (original product) and variation (duplicate). For a two-variation test, that's 50/50. You can adjust these percentages to any split you'd like — just make sure they add up to 100%. A common alternative is 70/30 if you want to limit exposure to the new version while still collecting enough data. You can also name each variation to make your reports easier to read (e.g., "Original" and "Bundle Offer"). ### Step 5: Choose Your Experiment Goal Select the primary metric that will determine your winning variation. Your options are: * **Revenue Per Visitor** — best for understanding overall revenue impact * **Conversion Rate** — best for measuring purchase likelihood * **Average Order Value** — best for testing pricing tiers and bundles * **Add-to-Cart Rate** — best for measuring initial purchase intent * **Checkout Start Rate** — best for measuring mid-funnel impact Elevate tracks all metrics regardless of which one you choose as the primary goal. Your selection determines how the statistical model identifies a winner. ### Step 6: Add Audience Targeting (Optional) By default, your experiment runs for all visitors. If you want to narrow the audience, you can add targeting rules to control exactly who sees the test. Common examples: * Run the test only for **mobile visitors** * Target visitors from a **specific country** * Limit to traffic from a **particular campaign** using UTM parameters * Target only **new visitors** or **returning visitors** You can combine multiple rules using AND/OR logic for precise targeting. For the full list of available targeting options and how to configure them, see [Audience Targeting](https://docs.elevateab.com/elevate-helpcenter/experiment-setup/audience-targeting). ### Step 7: Create and Review Once your setup is complete, click **Create Experiment**. Your experiment moves into a review phase where Elevate performs initial checks to ensure everything is configured correctly. ### Step 8: Copy Reviews to Duplicate Products If the original product has reviews, you'll want to transfer them to the duplicate so both variations show the same social proof. On the experiment review page, Elevate provides the specific information you need to send to your reviews app provider. They'll replicate the reviews onto the duplicate product. This keeps the experience consistent and prevents the duplicate from looking empty. ### Step 9: Sync Third-Party Integrations Make sure any third-party apps that affect the product page — subscription options, bulk purchase discounts, bundles, upsells — are set up on the duplicate product too. If a visitor lands on the duplicate and something is missing or broken, it'll skew your results and create a poor experience. ### Step 10: Submit for Elevate Review This step ensures the duplicate product behaves consistently across your entire store — no visitor will see unexpected or broken experiences in search results, collections, or recommendations. Click **Submit for Review** and the Elevate team handles the configuration. This is a one-time setup; once it's done, all future product tests skip this step. Review typically takes **2–12 hours**. ### Step 11: Quality Assurance Before launching, do a thorough QA check to verify both product pages work as intended. Check: * The duplicate product page loads and displays correctly * Reviews are showing on both products * Third-party apps (subscriptions, bundles, etc.) are working on the duplicate * All variation-specific changes (price, copy, images) display correctly ### Step 12: Launch Once QA is complete and everything looks good, click **Launch Experiment**. Your test is now live and traffic will begin splitting between variations. *** ### Next Steps * **Monitor your experiment** — Check the [Results tab](https://docs.elevateab.com/elevate-helpcenter/analyze/reports/reports-overview) on your dashboard to track performance as data comes in * **Wait for statistical significance** — Don't end your test early. Let it run until it reaches a definitive status. See [Statistical Significance](https://docs.elevateab.com/elevate-helpcenter/analyze/reports/statistical-significance) for how this works. * **Review advanced analytics** — Once you have enough data, use the [Advanced Analytics tabs](https://docs.elevateab.com/elevate-helpcenter/analyze/reports/reports-overview) to dig into visitor segments, checkout behavior, and per-product breakdowns