Understanding what A/B testing is and how it works is basically the easy part. It’s making sure that you don’t mess up when you’re actually running the tests that make up the trickier part of the process. One smallest mistake on your side would immediately make your results completely worthless and you’ll have to start all over again from scratch.
To prevent yourself from wasting your time, consider the list of dos and don’ts below when you start working on A/B testing your website.
(Courtesy of cocoate.com)
DO take as much time as you need
Quality of results – which means reliability and validity – is what’s important here and not spending the minimum amount of time running the tests.
Results aren’t valid if you’re not testing the right sections. Let’s say you want to know how good your articles are at drawing customers to your website. With that said, using metric tools to determine the age group of your market will absolutely make no sense because they have nothing to do about it.
Results aren’t reliable if you’re not using the right tools for measurement. Say you want to know how effective your landing page is in terms of sales conversion. The results won’t be reliable if you attempted to determine it just by counting the number of people who visited your website through your landing page. You need to know as well how many of those people actually bought something from your website.
And lastly, it goes without saying that for the results to be valid and reliable you should also spend the appropriate amount of time running those tests. A single hour will definitely not cut it!
DO NOT assume that A/B testing would always be sufficient
There will just be times when A/B testing is insufficient. If your website’s layout seems to be at fault as to why you are losing customers day by day, simply changing the header won’t cut it. You need to figure out as well if the color scheme or the use of Flash introduction is to blame. You then have to create different combinations with those changes you are proposing to make to improve your website layout. All in all, it means that what you need is multivariate testing and not A/B testing.
DO test both elements simultaneously.
You’re probably wondering how this is possible. Would you have to take down one version and put up another every minute? Isn’t that rather time-consuming or confusing to the visitor who spends more than a minute on your website?
Well, it is if you do it manually but you won’t have to do it manually. You can use various web tools that will allow you to conduct such tests simultaneously but without affecting a user’s overall experience when browsing your website.
DO NOT stop testing!
Performing A/B testing once and getting amazing results from it does not mean your job is over. Your website will have to continuously evolve if you want it to stay on top of its market and ensure that your customers keep coming back. For this reason, it’s critical to perform tests regularly and not just on the same section over and over. Run tests on other areas of your website to figure out which ones need changing.
DO use A/B testing for other marketing strategies you have outside your website
Think of email campaigns, e-zines, or even your blog. To give you a better idea of how to perform A/B testing on marketing strategies outside your website, let’s take your blog as an example. You can use the following areas for A/B testing and you’ll likely draw insightful results from your tests.
- Visual and writing style used for creating blog titles
- Length of blog posts
- Platform used for blogging (hosting)
- Use or absence of images in your blog post
Ruben Corbo is a freelance writer that writes about technology, gaming, music, and online marketing especially topics about A/B split testing and how to run multivariate tests. You can read more information on http://www.maxymiser.com for a better understanding on how it works. Ruben is also an avid gamer and music composer for short films and other visual arts.