Getting Started with A/B Testing

AB TestingA/B testing created quite a stir when it was first introduced as a viable testing method for optimizing websites or email copy. There is actually a lot to be excited about when it comes to A/B testing because it can get you actual results, even if you don’t have extensive experience in computer programming.

You might not be aware of it, but you’ve probably already visited sites that use A/B testing. You might have even participated in an A/B test already. A lot of big sites are already using this method to optimize their web pages and optimize conversion rates, such as Amazon, eBay, Google, Microsoft, and the BBC.

What A/B Testing Is

So what is A/B testing (also known as multivariate testing)? Basically, it’s a method where you present two versions of one element to your site visitors. For email marketers, it involves the variation of one element in the email copies that you send out to the subscribers in your list.

Generally, A/B testing only calls for the variation of only one element at a time. This is so that the results obtained can be directly and clearly attributed to the modification of that single element.

Variations of A/B Testing

Some people have come up with variations to A/B testing over the years. Instead of testing just one element at a time, several are modified per test. This way, a better design can be implemented by conducting fewer tests and within a shorter amount of time. The downside is that the reason for the results cannot be determined clearly, since it would be difficult to conclude which element caused the resulting success or failure of the design.

How to Conduct A/B Tests

  1. Make a list of possible test elements. In this stage of the process, it would be helpful to have the team work on putting together a list of elements that can be tweaked and modified. For websites, this could include the header, banner images, text formatting, font face and type, image placement, color theme, site logo, and navigation bar location.
  2. Select your test elements. You can simplify the process by ranking the elements you listed down in the first step according to urgency or priority. Once you’ve decided what element to test, you should move on to the next step.
  3. Set up your test. Create two versions of the element that you want to test out. For example, if you’ve chosen to test out banner images, you should create two versions of banners that are distinct and different from the other. Having too similar variations will not get you very good results.
  4. Conduct your test. For websites, you can install a script on your web server so that it will display the two versions of your website simultaneously and randomly to your site visitors. You can also use online tools like Google Webmaster Tools or hire providers to take care of conducting your A/B tests for you.
  5. Interpret the results. Which of the versions showed the best results? Which variation got more clicks, made more sales, or produced better results for you, based on your own metrics? Once you’ve determined which version performed better, it’s time to implement it on your actual website.
  6. Conduct more tests. Move on to the other elements that you’ve listed down in the previous steps. Website design optimization is an ongoing process, so conduct as many tests as it takes to come up with the best version of your website.

In conclusion, A/B testing is imperative from a business perspective. It allows you to determine the best setup for your website. As all successful businessmen know, it is important to test a hypothesis before implementing it. Using the proper steps greatly increases your chances of success. For an entrepreneur, A/B testing is that preparation for success.

EJ Parfitt

Tech writer EJ Parfitt has been writing for a short time now and has already picked up steam with several tech and sports sites . During his free time , you're sure to catch him competing in local chess tournaments in downtown Fort Lauderdale FL .

Latest posts by EJ Parfitt (see all)

Comments

  1. Thanks for nice tips .  A/B testing only calls for the variation of only one element at a time. This is so that the results obtained can be directly and clearly attributed to the modification of that single element.

  2. Glenn Verdult says:

    Hello i am Glenn Verdult I enjoy reading your articles I am looking forward to read more..

  3. Great tips!  Thanks for clarifying

  4. This is an interesting article. We look forward to seeing how this continues to develop.

  5. This is an interesting article. We look forward to seeing how this continues to develop.