subject: The A/b Split Testing [print this page] Conduct some simple a/b split testing to determine the best performing elements in your sales process. Perhaps you have an email campaign, or banner ads or an entire web site that needs tweaking. As a simple form of split testing, a/b split testing can help you see what works and what doesnt.
A/b split testing can be applied to many aspects of your campaign. In a nutshell, it means you choose one element from what you want to test, (say a heading from a web page), and then decide what you want to test about it (color, font size, content). You can only test one thing at a time to ensure accurate results. Otherwise, you might not know for sure what caused the better yield.
What you do is create two versions of that heading: version a and version b. Then you set up your server to alternate the versions between visitors to your site. For example: visitor one sees version a, visitor two sees version b, visitor three sees version a etc. At the end of your test period, whichever version give you better conversion rates is kept as is, or retested against another version. Dont stop there, though. Keep a/b split testing all parts of your web page, because little things can make a big difference. Even something as unobtrusive as a background color could me the difference of getting a click through or not getting one.
As for deciding how to break down your list or your visitors for the purpose of a/b split testing, it really is up to you. If you have heavy traffic to your site, then you can probably afford to run a larger test, if not, you might want to run a smaller one. You just have to make sure its large enough to give you accurate results. If youre testing a web site, you can tell the server how to show which version. If you are conducting email a/b split testing, you can send them yourself.
Although a/b split testing usually brings ideas of creating two new versions, it doesnt necessarily have to work that way. If the element you are testing seems to be working pretty well, and youre just trying to see if it can be better, you might want to just create one new version, and test it on a small percentage of your traffic while still running the old one on the majority.
A/b split testing can be used to test emails, landing pages, other web pages, and ads of all kinds. You can test subject lines, headings, font size and colors, background colors, graphics, graphics vs. text, links, call to action buttons, ad size, ad placement, and content. If its purpose is getting customers, it can benefit from a/b split testing.
Although a/b split testing is relatively simple to do on your own, if you need help, many tools exist in the internet, for that purpose. Some are free, some are not. Conduct some research. Whether you decide to use a tool, or do it on your own, a/b split testing can help you improve your result.
