Shopify
Cloudflare Workers
Klaviyo
Slack
Google Analytics 4
Microsoft ClarityPath to purchase
Funnel per group
Where each group's visitors leave, step by step.
Test results
Know when a result is real
Conversion, revenue and profit per visitor from Shopify's own orders, with a data-quality check before any winner is named.
Product shot beats the current hero
Data looks sound · 92/100Revenue per visitor up 8.4% (95% interval +2.1% to +14.9%). Chance to beat the current hero: 97%.
Chance to beat control97%
Conversion rate
+5.2%
−1.8% TO +12.6% AT 95%
3.10%
3.26%
CurrentProduct shot
Revenue per visitor
+8.4%
+2.1% TO +14.9% AT 95%
$2.37
$2.57
CurrentProduct shot
How results are read
Numbers you can defend in a meeting
Orders from Shopify
An order counts for a group only when Shopify's record carries the visitor's cookie or confirms a pixel claim for the same order.
TraditionallyBrowser-reported purchases
With Growth Dept CROShopify's own orders
Per visitor
Every metric is worked out from each visitor's own row, with Holm's correction when a test has three or more groups.
TraditionallyRates averaged per day
With Growth Dept CROOne row per visitor
Gated
A sample ratio check and a data-quality gate run before a winner is shown, so a broken split cannot crown a group.
TraditionallyA winner at 95%, every time
With Growth Dept CROA winner when the data holds
A test result is a claim about your store. Growth Dept CRO only makes it when the orders, the split and the sample all hold up, and the verdict says so in plain words.
How the verdict is written
Every view a result needs
Key metrics
Up to seven key metrics per test
Conversion rate, revenue, profit, AOV, discounts, shipping and subscriptions, each with its interval and chance to beat the control.
Statistics
The checks behind the verdict
Sequential monitoring, power, CUPED, a bootstrap check, novelty and drift, attribution windows and guardrails, in their own views.
Act on it
Roll out, share or export
Roll the winner out, share a read-only link that expires, export the totals or the line items, and post the result to Slack.
Views
Cut every result the way you need it
Order sizes
Order distribution
How each group's order values spread.
Roll out
Keep the winner
To everyone, an audience or a link.
Over time
Values over time
Every metric by hour, day or week.
FAQ
Questions, answered
Do you drop outliers?
Optionally. Large orders can be capped at a percentile of buyers' revenue, 99th by default. It is off by default so the default read matches the engine's tested numbers.
Can I share results with someone outside the team?
Yes, with a read-only link that lasts 7, 30 or 90 days and can be revoked. The shared page shows the results views with the filters fixed.
Find out what is worth testing on your store.
Book a 45-minute call with the Growth Dept team. We look at your store and your traffic, and tell you which tests to run first.