Blogs

Testing elements of your mailings is one of the most effective ways to improve performance of your marketing campaigns and with email, unlike off line media (print, radio, TV) there is almost no additional cost associated with deployment.

When I talk to customers and prospects, I often ask if they are currently doing or planning to do any A/B testing of their messages. I am always surprised when they say that they want to, but don't. Most of them cite time and effort as the main reason, but some also say that it's too difficult to set up the data and track everything.

Pishaw!

A/B testing your email program doesn't have to be a cumbersome, laborious task. If you aren't doing any testing, it's easy to just try one or two things to see if they have any effect. And if you are thinking that it requires re-working your content, there are other aspects of your mailings that you can test that don't require any effort from your creative team.

In Message Studio 4, doing A/B testing is a piece of cake. Message Studio takes care of splitting and randomizing the data and makes it simple to configure each test with the variables you want to evaluate.

You can create between 2 and 20 splits in a single mailing, though if you are just starting out with A/B testing, I recommend that you start with a small number of splits, say 2-3. Within each split you can control a variety of variables including:

1. Template
2. Format (HTML, Text or multi-part)
3. Headers (From Name, From Address, Subject, etc)
4. Launch date and time

This is a great way to try out new elements of your mailing. The A/B testing feature in Message Studio lets you send the test to all or part of your subscriber list, so you can test a particular change with just a sample of your list (e.g., 20%) without affecting everyone else. After you send the tests and determine the winner, you can easily assign those parameters to the remaining unsent recipients and then launch. There's no need to create another mailing or figure out who received the test email and who didn't.

A/B testing doesn't have to be a one-time process either. You can test and tweak as much as you like. Once a month, week or day. Each time you do, you get valuable information about your subscribers and what they respond to in your emails.

Once you have mastered testing single mailing elements, you can move on to multivariate testing to learn how you can fine-tune your email campaigns even further.

For more details on how to configure an A/B split mailing in Message
Studio, check out pages 129-130 in the Message Studio 4.1 User Guide.