What is A/B Testing?
An A/B test is an optimization tool that allows marketing teams to understand what elements of a website work best. They do so by creating two variants, variant A which is the original website outline, and variant B, which is a very similar adaptation off the same website but with small changes made. By testing both variants, we are able to see if those small changes had any effect on meeting our goals more effectively. For example, a small difference in coloring may result in more clicks, a common goal websites have.
Creating a Hypothesis
Before just going and changing everything on the website, it is important to first “identify a problem”. After, you must come up with an idea on “what you can change to improve it” https://support.google.com/optimize/answer/6211930#create-a-hypothesis
In other words, you must create a hypothesis. An example of this process would be identifying that your video is not reaching your goal of video clicks. This is the problem.
With this problem in mind, you can then create a hypothesis. For example, the hypothesis could be that if you move the video to the beginning of the page rather than the end where it was originally placed, you can get more clicks on it. This would be the hypothesis to the problem of not meeting our video click goal. We can then test this hypothesis. The video placed at the end of the page will be our variant A. The website with the video placed at the beginning will be our Variant B. After deciding who and when to target and actually running the experiment, you will get a lot of new data. This new data from both websites will help to visualize if the hypothesis was correct and the website is better optimized with the video at the beginning.