Introduction
An overview of Elevate A/B Testing for Shopify merchants: what you can experiment on, how the Liquid snippet and Web Pixel deliver and track experiments, and where to go next.
Elevate A/B Testing is an A/B testing and personalization platform built for Shopify. It lets you run controlled experiments on your store, testing different prices, product offers, page layouts, images, shipping rates, checkout elements, and themes, so you can make decisions from data instead of guesswork.
Everything runs on your live store. Elevate splits your traffic, tracks what each group does, and tells you which version performed better.
What Can You Test?
Find what you want to test, then open the experiment type that handles it.
Different price points on your products
Page templates and layouts
Two entirely separate pages, at different URLs
A different product offer: bundles, descriptions, images, or a full alternative version
Your whole theme, across the entire store
Shipping rates, thresholds, and delivery options
Product photos and the order they appear in
Checkout banners, bundles, testimonials, and trust badges (Shopify Plus)
Text, images, and styles, using our content editor. No code required.
Anything else, using your own JavaScript and CSS
Personalizations
An experiment splits traffic to find a winner. A Personalization does something different. It delivers a specific experience to a specific audience, continuously, with no winner to declare. Use it when you already know what a segment should see: member pricing, a region-specific landing page, or a campaign offer for paid traffic.
Prices, for a specific audience
Advanced Pricing Personalization
Page layouts, banners, or content
Page Personalization
Product titles, descriptions, or layouts
Product Personalization
Where an audience lands
URL Redirects
Which product images an audience sees
Product Image Personalization
Content, images, or layout, with no code
Content Personalization
Anything else, using your own CSS or JavaScript
Custom Code Personalization
Checkout messaging and offers
Checkout Personalization
How It Works
Create an experiment. Choose your experiment type, set up your variations, and define your goal metric.
Allocate traffic. Decide how much traffic each variation receives. By default, traffic is split evenly.
Launch. Elevate serves the variations and tracks visitor behavior automatically.
Analyze. Monitor results in real time. Elevate's Bayesian statistical model tells you when a winner is identified.
Act. Apply the winning variation to your store, or roll it out as a Personalization for a specific audience.
What Elevate handles for you
You enable the Elevate app embed in your theme. There is no code to write and no snippet to paste.
From there, Elevate manages delivery through a Liquid snippet, tracks visitor behavior through a Web Pixel, and evaluates results with its statistical engine.
Does Elevate duplicate my products or pages?
For most experiment types, no.
Theme Experiment, Page Experiment, Product Experiment
Price, Content, Custom Code, Split URL, Shipping, Product Image, and Checkout Experiments
Where a duplicate is needed, Elevate creates and manages it for you, and removes it when the experiment ends. Price Experiments no longer require duplicate products on any Shopify plan.
Where to start
New to Elevate A/B Testing? Install the app, verify your installation, then create your first experiment.
Know what you want to test? Go straight to the experiment type in the table above.
Need help reading results? See Reports Overview to understand your experiment data, or Statistical Significance to know when to trust it.
Have a question? Check the FAQ or contact support.
Key concepts
These apply across every experiment type:
Experiment Lifecycle: How an experiment moves through Draft, Preview, Running, Paused, and Done
Audience Targeting: Control which visitors are eligible for your experiment
Traffic Allocation: Set how much traffic each variation receives
Traffic Isolation: Reserve a share of your traffic so those visitors join only this experiment. Off by default.
Product Costs: Add cost data so Elevate can report profit alongside revenue
Statistical Significance: How Elevate determines a winner
Ending an Experiment: What to do when your experiment is complete
Last updated