As a business owner, one of your main goals is figuring out what your customers want and need. The art of deciphering what your customers want is a tricky and sometimes impossible task. The list of ways you can get feedback from your existing customers is almost endless. However, how do you figure out the needs of people who are not your customers yet? Web developers pondered over this dilemma and developed A/B split testing – a website optimization technique designed to target potential customers.
The Basics of A/B Split Testing
An A/B test is quite simply a test with two options - A and B. A/B split testing in website development is a little more complex, but the basics still apply. Any A/B split test begins with two versions of your web page, where each version has slight changes. Some visitors will be shown version A. Others will be shown version B. The results of showing visitors one version versus another are then tracked to see which provides the more desired outcome.
Like any experiment, when beginning A/B split testing it is important to set out with a hypothesis in mind, that way your A/B split testing does not turn into A/B guessing. “More desired outcomes” can mean anything you want them to - and they are pretty pointless if you do not know what you are looking for.
What to Test
Because the scope of A/B split testing is so vast, many people are either unsure where to start or unsure where they are going. As a result, test results are misleading or just plain wrong. Your best bet is to start with simple changes. Many people when overhauling their website first start with calls to action. This means anything that will cause your visitor to take immediate action. Examples of these on your website might be “Buy Now,” “Like Us,” or “Subscribe to This Blog.” Calls to Action are the easiest to test because you have definite results - did the respond to the call of action or did they not.
Another great A/B split test to perform is with your sales funnel. Were you able to get people from your blog to your Services page? Were you able to get people from their virtual shopping cart to the virtual check out? Funnels are another easy change to test with bounce rates and exit rates, once again providing you with a clear answer.
The copy on your webpage and your visuals are two other great factors to experiment with. These two are slightly more difficult to test because they do not have a black and white outcome, but they can still prove just as helpful, given that you test with a hypothesis in mind. Do you think changing your copy will keep viewers on your page longer or direct them to your “Contact Us” page quicker? Do people prefer to see another person on your website or an animal? Does changing the photo bring a different demographic to your website?
Experiment with Caution
At first read, A/B split testing sounds too good to be true. And for those who do not approach it cautiously and critically, that ends up being true. Make sure to check back for the Top 5 Mistakes of A/B split testing and then feel free to experiment away, you mad scientist you!
