• ANALYTICS AND A/B TESTING • SETUP GUIDE

How to Compare A/B Test Results

By Rivio Support Team
1 min read

What you will achieve in this guide

Follow this step-by-step documentation to configure custom B2B registration forms, approval workflows, and customer tagging.

EXPECTED SETUP TIME 5-10 Minutes
COMPATIBILITY All Shopify Themes & Plans
TECHNICAL DIFFICULTY Zero Code Required

Compare registration form variant views, submissions, conversion, and the app’s significance check before choosing a design.

Before you begin

Select the form and reporting period you want to inspect. Use a known storefront visit and application as a test, and keep visit, submission, approval, and sales counts separate when interpreting results.

Compare rate, sample size, and significance

Each variant has recorded views, submissions, and conversion. A higher conversion percentage based on a small sample can be unstable, so read the app's significance information alongside the counts.

Keep the variants unchanged during the comparison. Changing a color or allocation midway means the whole reporting window no longer represents one consistent experiment. Demo traffic can prove setup works but is usually insufficient to establish a winner.

Example

Consider an illustrative result of A: 100 views, 10 submissions and B: 100 views, 15 submissions. Compare 10% with 15%, then inspect sample size and the app's significance result.

Expected result

The higher observed rate is not automatically a proven winner; demo numbers are illustrative rather than a prediction.

Limits and important behavior

  • Small samples can produce unstable conversion estimates.
  • Check that both variants received traffic before declaring a winner.
  • Do not change multiple variants repeatedly and treat the entire period as one unchanged experiment.

Troubleshooting and verification

If a variant appears to win after only a few visits, compare sample sizes and the significance indicator before deciding. Keep the same period and form across variants. A visitor keeps the assigned variant, so repeatedly refreshing one browser is not an independent sample.

Screenshots to add

Screenshot placeholder: Capture the app screen or test result described in “How to Compare A/B Test Results”. Include the relevant saved settings and the example outcome so readers can connect the explanation to the interface. Use dummy applicant details and mask any secrets.

Related guides