Split Testing Email Newsletters

Here’s something you die-hard online direct marketers should love – Aweber has a facility to do split testing of broadcast emails you send out. I just sent out an email to my Entrepreneur’s Journey email list (you join by filling in the form above next to the big red subscribe arrow) and I noticed the split-testing option for the first time. I haven’t used split testing yet but seeing how excited some of my marketing friends get when it comes to split testing I know it’s a powerful tool.

What Is Split Testing?

In a nutshell split testing allows you to expose multiple elements to your audience and test them to see which works better. In most cases it’s all about increasing conversion rates. Taking an email broadcast example, if you are going to send out a broadcast email to your mailing list and that email is designed to elicit a response from your audience, you can use the split testing option to send out up to four different emails and see which gets the best response.

This is how Aweber describes the feature -

You may test up to 4 split test groups of varying percentages. For example, if you want a split test with 10%, 10%, 10%, and a final group of 70% it will automatically randomly divide your opt-in list into unique groups. You can then enter your broadcast text into those messages and send them. Send your broadcasts all at once or wait for feedback and results before mailing each group. If you want to create less than 4 groups just enter the percentages for those groups. For example, 25%, 25%, and 50% would great 3 groups.

Pretty cool hey! If your business depends on email lists, and most online business in some way or the other do, this is a fantastic tool that provides the analytical marketer exactly what they like – statistics on conversion rates and a means to tightly control the split testing process.

Keep Your Variables Controlled

The important thing with split testing is to be certain to test each element independently. If you change multiple elements in each test then you don’t have any clear relationship to correlate a change in response, it could be because of any element change.

For example if you send out an email broadcast to your newsletter and you want to increase the number of people clicking a link in the newsletter you can split test the anchor text (link description) in different email broadcasts. If you hold all the other elements constant and send out emails with only different anchor text you can learn which generates a better click-through rate. This information can then be used and refined for future broadcasts.

Statistics Don’t Lie

Marketers love split testing because statistics don’t lie and have no emotions tied to them. It’s simply about the numbers, in particular the conversion rates. While no split test can be 100% accurate because we are dealing with humans, you can draw some fairly conclusive statistics provided your sample is large enough.

Tools like Aweber make this process completely automated providing you with easy to use forms and clear basic numbers at the end. Once your split test is complete you have percentages explaining the results of your test and provided you kept your variables in control your results will be statistically valid.

Yaro Starak

PS – I actually hate statistics – I even failed business stats at university, but I’m very aware of how vital they are to business and how easily accessible they are to online business owners. I’d be stupid not to take advantage of them especially when when we are talking about things like split testing, which is not too complicated.