A/B Split Testing for Determining Webpage Effectiveness

Posted on Monday, November 25th, 2013 at 1:28 pm by shauli

A/B split testing is effective in comparing the effectiveness of two different versions of a webpage.  This is a great tool to utilize in order to implement the website with the largest rate of responses.  When using A/B split testing, you will always have to choose whether to stick with the webpage you are currently using, or to switch to the alternative site.  If, after you have completed the A/B testing, there is a significant difference between the two pages, the decision will be an easy one.  Obviously, you would pick the better of the two versions to increase traffic on your page.  The decision to switch to a new site when the difference is small may be harder to decide, but since the difference is small, that decision will most likely be insignificant.

What happens, though, when the difference between the “control site” and the new, “challenger site” is large, but the decision is still unclear.  Sometimes the data found doing these tests may be hard to read or difficult to understand.  One example is with the use of “sign up for a free trial” boxes located on a webpage.  The “control site” has the box on a side panel as well as a bit farther down on the page.  The concern is that although the second box may disrupt the flow of the webpage, it may produce more results since there are two options to sign up for a free trial.  A/B split testing is a perfect way to decide whether this second box is necessary.  With the “control site” having the two free trial box options and the “challenger site” without, an A/B split test will deem the effectiveness of one or two boxes.

After a 14 day trial, it was concluded that without the second “free trial” box, there were 15.6% less sign ups.  However, at the same time, sales increased by 1.5%.  The average purchase and total profits made on the site was also higher when the second sign up box was removed.  It is hard to decide whether to forfeit the increase in sign-ups for the bump in profits or the other way around.  A/B split testing is helpful, but make sure you weigh all of your options before deciding to switch up your website.

Leave a Reply