A/B Test Founded Decision Making Basis Instead Of Guessing Game
Management Summary
There are several common options that can lead to a decision: a vote is often carried out among colleagues or the boss is ultimately allowed to decide. With such a decision-making basis, whether your potential customers would have converted more if a version of your campaign had been designed differently will forever remain a secret.
Therefore, a few questions arise: On what basis is the decision made? At the end of the day, which version will be better? Drumroll…and the answer is:test, test, test. Because without testing it can never be claimed that one of several variants is definitely better.
A/B test – the mechanism
In an A/B test (also called a split test), two versions of the same page are tested against each other. 50% of visitors see the original version, 50% see a modified version. Quantitative data doesn’t lie: The version that converts significantly better can undoubtedly be chosen as the winning variant and installed on the website.
Case: A/B test identifies the better variant of the OptIn Layer for bauMax
Using the example of bauMax, we show that by testing two different newsletter opt-in layers, significant conversion rate increases can be achieved. Two variants were tested that only differ based on their color: yellow vs. red.
First, a Opt-In Layer rule mechanism must be definedso that the layer is only displayed if certain criteria are met. Each variant is then played out to half of the website visitors. In the case of bauMax, the winning version was the version that generated more newsletter registrations than the other version:
A short time for you to ponder…Which option was better? Yellow or red? If you guessed the red variant, you are wrong. The yellow opt-in layer achieved a highly significant uplift of 14.8% compared to the red variant and thus a better conversion rate.
Without testing, the decision might have been made in favor of the red variant from the start and thus the worse-performing variant would have been played. Many newsletter opt-ins would have been missed by bauMax if it had made the wrong decision without testing.
