A/B Test
Definition
An A/B test shows two versions of a page, element, or offer to different visitors at random and measures which one converts better. It replaces opinion with evidence for decisions such as button text, product page layout, or how a price is presented. A test needs enough traffic and one fixed success metric, or the result is noise.
Why it matters for your store
Everyone on a team has an opinion about the product page. An A/B test settles it with data, and it protects the store from changes that feel like improvements but lower conversion.
The method has rules. Split visitors at random and keep each visitor in the same version. Decide the metric before you start: conversion rate, average order value, or revenue per visitor. Run the test long enough to cover weekday and weekend behaviour. Stop only when the difference is large enough to be more than chance. Testing one change at a time makes the result readable; testing five at once tells you nothing.
Small stores rarely have the traffic for fine-grained tests. They do better testing big, structural changes, and reading the result over a longer period.
How Kambloo handles it
Kambloo's checkout is built so that a single-page layout and a step-by-step layout render from the same data, which is what makes a fair comparison between them possible at all.
Until then, the campaign performance and coupon reports let you compare offers over time. That is weaker than a controlled test, because periods differ, but it is enough to see whether a threshold or a discount changed order value.
Questions about A/B Test
How much traffic does an A/B test need?
Enough conversions in each version to tell a real difference from chance, usually hundreds per variant. A store with a few orders a day should test big changes, such as a checkout layout, not button colors.
What should I test first?
The steps with the largest drop-off: the product page to add-to-cart, and the checkout. Changes there move revenue; changes on the about page do not.