A/B testing and the Right Creative
Thar She Blows
Relying on A/B testing to turn out the right creative can be about is elusive as captain Ahab’s search for the white whale. At times it can feel like you are approaching a multifaceted problem with only a pair of pliers in your hand. Don’t get overwhelmed, just take pause and remember, Rome wasn’t built in a day. Then remember, you are not trying to build Rome, your goal is to build something that generates revenue for you.
A/B testing is a great way to do just that. It is simple, effective, and non biased (which can be difficult to swallow at times when your intuition tells you something else). The good news is you have a site that has traffic coming to it already. The better news is you have a tool to help you convert that traffic into sales. All you have to do is choose a creative element and get started.
There are several different creatives you could test, but a couple of good ones are your email campaigns and your landing page(s). Why? Well simply put they are generally the two largest prongs of your sales force. They work on basic in and out principles. Your email campaigns go out to your customers. Your landing page is designed to bring customers in. Each works differently, yet utilizing A/B testing makes them much more effective.
Once you decide which creative to use, you must then decide which element to test. If it is an email newsletter, you might want to test the placement of your call to action button, or certain color schemes. All of these elements can be tweaked to turn traffic into customers. Make sure you are running a good analytics program so you can easily go over your data and make decisions later.
Proper A/B Testing
Finally, there are a few things you need to consider when it comes to A/B testing. These are probably not new, but it never hurts to keep them in the back of your mind. First, you always want to make sure you test both the control group and the variable group at the same time. Splitting traffic between these two elements will keep it steady and unbiased. If you test one before the other you have introduced another variable. Maybe the traffic increased because of the time of the month or a certain holiday. Testing both at the same time eliminates that.
That last point is actually the second point and bears repeating. It is important to keep in mind that with A/B testing you are only testing one variable. That is why it is called A/B testing. If you test two or three variations of on a product page or newsletter at once, you really have no idea what makes the difference in traffic increase. Sometimes it is item placement, sometimes color, other times it might be the type of font you choose. In order to know decisively you must only test one variable at a time.
Finally, keep an open mind. A/B testing can yield some surprising results at times. It might go against what you are thinking or what would traditionally be expected. That is ok. The goal is to generate revenue. Trust the results. Trust your new creative element. Put it to work and let it generate revenue for you!