A/B- and Multivariate Tests.

The A/B-Test is a test method for evaluating two variants of an eCommerce system, in which the original version is tested against a modified version. The aim is to increase a certain user action or reaction. The A/B test divides the target group into two subgroups: Group A and group B. The test object is also divided in two according to the distribution of the target group: the original variant and a changed variant. The original version is then used for group A and the modified version for group B and the reactions are compared. Reaction refers to the desired effect, such as registering, subscribing to a newsletter or ordering a product. In contrast to the multivariate test, the A/B test only changes one variable and tests its effectiveness. For an A/B test to be effective and the results to be valid, a sufficient group size must be given.

With A/B-Tests you make decisions based on reliably measured behavioral data of your users. We do not measure opinions, nor intentions. We use A/B tests to measure and optimise click behaviour, purchasing behaviour, registrations or registrations. The test variants are tested against each other in live operation. This gives them data and insights based on actual behavior. The results from A/B tests are valid and show you reliably which test variant is the best in terms of your key figures. With the Smart Commerce A/B tests you get the security to optimize the right things correctly!