Customers and loyalty
Two strategies.
One test to understand.
Similar products or complementary items? An A/B test compares two alternatives against the same goal. Choose a scenario, generate simulated visits and orders, and observe the comparison.
Two ways to recommend.
The one that sells wins.
On the same backpack’s product page, half of the visitors see strategy A and half see strategy B. Start the comparison and watch simulated visits and orders arrive: the demo only declares a winner when the data allows it.
Other backpacks in the same category and price range.
Products that people who buy this backpack often add to their order.
How to read the result
- Conversion is the number of orders divided by the visitors assigned to that strategy, including those who clicked nothing.
- The coloured band is the uncertainty: wide with few visitors, narrower with many.
- A winner is declared only at the end of the planned sample, and only if the interval of the difference does not include zero.
Demonstration with fictional data: it runs entirely in this page and does not represent the results of a real store.
What to look at
Start with the visitors assigned to each variant. Then compare orders, conversion rate and revenue per visitor: each describes a different aspect of the test.
Try a scenario with no difference
Small samples can make a variant look better by chance. The bands show uncertainty and the simulation gives its verdict at the end of the sample. These fictional results do not predict your store’s performance.
See it working
on your catalogue.
We’ll show you the storefront and the back office with products like yours, and how you would get there from your current store.