subject: How To Conduct A/b Split Testing [print this page] How To Conduct an A/B Split Test How To Conduct an A/B Split Test
If youre wondering how well your campaign is performing, an a/b split test can give you some answers. A form of split testing, you can use it to test any part of your campaign or sales process. If it can affect your conversion rates, it can be split tested, and the simplest way to do that is to conduct an a/b split test.
If you are unfamiliar with split testing, the a/b split test is the best place to start. Basically, it means to create different versions of whatever youre testing, and split them among your potential customers to see which ones result in the most sales. Because of the nature of the a/b split test however, you can only test two versions (a and b) of one element at a time.
It doesnt matter what kind of marketing platform youre using; split testing in sales began in direct mail campaigns, now it has branched out into eBusiness, it can be improved by an a/b split test. It works for testing emails, landing pages, all types of online ads, web pages, etc. You can test subject lines, headings, links, graphics, fonts, sizes, colors and placements. You would be amazed at the big returns a small change can bring.
Well cover some a/b split test basics by using the simplest example: emails. The idea behind email marketing is to not only get the recipients to open it instead of just treating it like junk, but also to read it, and follow through on what you want them to do. Before that can happen however, they do need to open it. How do we get them to do that? With a catchy subject line. How do find the best subject line? By conducting an a/b split test!
Come up with a few subject lines that you think will get a good response. Then chose two for your initial a/b split test. Although other types of split testing allow you to test more variations of more elements, with the a/b split test, you can only test two variables of one element at a time, else you wont be able to tell which version of which element made the difference.
Create two emails that are identical in every way except for the subject lines you chose. Then you need to decide what kind of sampling you want to test. Do you want to risk half of your mailing list on a poorly performing subject line? What if both are bad, then you might have a serious problem. Anyway, chose a percentage of your list that is large enough to give you a valid result, but not so large as to create a negative impact on your business. Of that percentage, decide how you want to split the two subject lines. Most like you would choose a fifty/fifty split of your sampling, but you dont have to. Send the test emails out accordingly, and wait to see what happens. The one performing best is the winner. Keep repeating the process until you have optimized that subject line, then use the a/b split test to optimize your entire email campaign.
