5 reasons to A/B test your website

5 reasons to A/B test your website

E-commerce currently represents 9% of the general commerce’s turnover (excluding food). With such an important rate, the efficiency of e-commerce websites is no longer an added bonus. It is a necessity. It is even more crucial as websites have an impact on the brand’s vision that consumers carry out. Issues that customers meet on a brand’s website are issues about the brand itself. Launching a new version of your website or pages without testing it might be catastrophic. If the new design and ergonomic does not match with your users expectation, then it can have a very bad impact on your conversion.

Here are the reasons why A/B testing your site redesign is necessary:

  • First of all, A/B testing justifies your choices. No longer will marketing decisions be based on intuitions or the feeling that your web users would like that new website version better. With A/B testing, your decisions are based on a true analysis of your Web users’ interaction with the different versions. A/B tests are a guarantee to select the version that will convert the most and avoid the underperforming choice.
  • A/B testing does not only help you to choose the version that converts the most but also helps you to choose a version that matches with your main objectives. It helps you to divide your audience and to do a segmentation in order to give the right version to the right person. The comparison of different website versions does not automatically reveal a “good” and a “bad” version. Version A can perform better in one objective or with a specific segment of Web users while version B performs better in other objectives and with another specific segment. The precision of A/B test allows you to determine where and with which type of user exactly each version performs better. With an A/B test, you might realize that version A is more efficient on the desktop while version B is more efficient on mobile devices.
  • You do not always have to choose between two versions that performs better with a different segmentation of users. Once the A/B test confirms the preferences of your users, you can simply decide to display the version A of your website to the user type that would more likely prefer it to version B, and so on. When well built, this personalization function has a huge positive impact on conversion.
  • Segmentation functions also represent another asset. During a first period after a website redesign, the conversion rates usually decrease because Web users familiar with the previous version lost their points of reference. By displaying the new version first to new visitors, and it incrementally familiarizes users. A/B tests avoid the risk of losing loyal customers.
  • A/B testing does not only help to choose the best version; it helps to think of the best version. Pointing out the elements that slow down the conversion on each version of your website associated with A/B tests can lead to an optimized version composed of every converting element from each of your tested versions.
AB test

AB test representation


In conclusion, A/B testing does not only avoid digital catastrophes, but also it leads you to optimize even more your solutions and achieve your objectives with the best versions of your website. A new website launch is a investment, which impacts the conversion rate and the brand’s image. Your solutions can easily be optimized when tested, but once the website is launched and the users lost, it is difficult to catch up.


Authored by: Marine Hortemel

Head of Marketing & Communication