There are a group of professional marketers out there that believe strongly that tracking and measuring the ROI on your marketing money is one of the best ways to optimize your spend.  They also believe that testing your marketing campaigns is one way to gain the highest possible return.

I’m one of those types of marketing professionals — and we apply that philosophy to email marketing where the tools and statistics are readily available to help you test your way to a better ROI.

Yet few people run A/B split tests regularly.  Sometimes they don’t know how (I’ll tell you how in a second) or sometimes they don’t know what to test (I’ll give you ten things to test in a second).

How to Run Split Tests

There are two main ways to run A/B Split tests.

First. You can split your list in two groups.  If your software allows a split test, they will give you a random split of A/B.  If not, you’ll have to create a custom field names “Split” and populate it with a ‘A’ value and a ‘B’ value.

Now you’ll have an ‘A’ segment and a ‘B’ segment.  Simply send one version to the ‘A’ segment and your alternate version to the ‘B’ segment.

Compare the results and voila, you’re split testing.  Over time you will learn how your audience responds and it will guide your email marketing decisions in the future.

Alternately, if your list is big enough.  You can take two test groups of 10K names each.  Send your ‘A’ version to one group and your ‘B’ version to the other group.  Compare the results and you could have a clear winner.  Even a marginal winner helps you.

Then release the winning version to your full list and take the benefits of the best performing message.  This only works if your list is large enough.  You could do it with 500 names in each test but the stats could mislead you.  (A statistician can tell you why)

Now on to the fun part.

What the heck can you test?  Here are 10 things to test in your emails.

1.  Subject lines – make them completely different or change one word.

2.  Body copy – write two different messages.  See which compells people the best.

3.  Body copy length – settle the battle of long copy vs short copy once and for all on your own audience.

4.  Email layout – test two columns against one column, big header vs small header, alternate formats.  Make your designer crazy.

5.  Time and Day of the Week – compare 6AM to 6PM on Tuesday.  Compare Tuesday AM to Sunday AM.  Don’t assume anything.

6.  Segments – test your target audience and see if you can zone in on a winning group. Better targeting equals better results.

7.  Your Call to Action – test text verses buttons.  Test location and frequency of buttons. Compare alternate text in your link.

8.  Testimonials – add testimonials to the message and look for a lift.

9.  Personalization – go beyond Dear Mitch.  Test personalizing other content in the message itself.

10. Images – test with and without images (you might be surprised) and test competing images.

If you just take ideas from this list, you’ve got 6 months of ideas to run A/B split tests to make improvements.  The biggest difference I’ve seen split testing make is a 12X difference.  That can be HUGE!

Enjoy your testing and let us know if we can help.