How to do A/B testing

Excerpted from ConvUrgency.com's Smart Email Marketing Guide

One of my favorite components of email marketing is the testing. I think that’s because it’s one of the most dynamic components. What I mean is this: you generally know what your offer or promotion will be. You’re steadily growing your list, and you usually settle on an effective but more-or-less standard email template. Testing however, can be done on so many variables, and will constantly yield different results.

Here are just a few of the variables you can test:

• different subject lines
• sender – company vs. your name
• ad copy
• image placement
• templates, colors, fonts
• action text (e.g. “order now” vs. “Hurry, offer ends soon!”)
• landing pages & conversion forms

And there are many more no doubt that come to your mind. But how do you test? How do you know which variable is best when each email campaign differs somewhat?

The easiest method is A/B testing, with a 10/10/80 split. A/B testing indicates that you’ve made 2 very similar emails, but with one variable different in either version – giving you an A and a B version. You send each version to a 10% segment of your list. After a few hours, or maybe a day, you see which version performed better, and send it to the remaining 80% of your list.

It’s really that simple, and you can refine this an endless number of times, continuing to tweak one element at a time until you achiever optimal open and click rates.

Another variable to test is send time. Over time, try sending your email on different days and times. You will find that some days delivers better open rates depending on your audience/content. For example, a hobbyist newsletter may be opened most on the weekend, while a B2B email more often during the week while people are at work. Find the best day/time to send, and stick to it.

Testing takes time, but it will enable you to get more out of each email you send.

Excerpted from ConvUrgency.com's Smart Email Marketing Guide