If You Want To Send the Best – Split Test!


According to our 2010 Email Marketing Trends, only 25.9% of respondents plan to use split testing to reduce the negative impact of inbox clutter. Now we know they don’t speak for you, but that metric worried us. So we decided to try and boost the numbers with a little more information!

Here’s one great reason to Split Test more often: Based on our recent study of GetResponse emails, we all know that subject lines matter, and personalization works. In fact, emails with personalized subject lines averaged 26% higher open rates and over 130% higher CTRs than emails without personalized subject lines.


But we also know that once you get beyond the obvious first name, thinking of something short for the subject line that will make every person open your email is NOT easy!  And we also heard that you’re going to increase segmentation and email targeting, which is great, but requires even MORE personalization.

So why not come up with your best efforts, and let GetResponse try them out for you?! It doesn’t take long, and it certainly reduces the chances of your wonderful campaign being tossed into the junk folder, or worse, flagged as SPAM.  Don’t risk it!


Speaking of spam, here’s another good reason to Split Test: Those dreaded SPAM words are changing all the time, so unless you’re checking in daily, a good way to see if your subject lines will pass the filters is to Split Test! Don’t risk you reputation simply because you think you have to hurry and move on to your next marketing task.


Think about how much in lost sales and sign ups are at stake!

Okay, enough lecturing, but before we go into the logistics and benefits of finding out which versions of your email will perform best – before you send − let’s review a few Split Testing Basics.



What is Split Testing?

Split Testing, or A/B testing as some call it, is the testing of different versions of your email to see which one delivers the best results −typically opens, click throughs, and sales or purchases. GetResponse customers can actually test up to 5 versions each of a message variable, however you can test as many or as few variables as you think is necessary:

  • Subject lines (impacts open rates!)
  • Headlines
  • From fields (company and/or names)
  • Body copy content (message)
  • Body copy length
  • Template design/format
  • Media, e.g. audio-video, photos, graphics etc.
  • Text vs buttons for calls to action
  • Type and number of products offered
  • Price points, discounts, etc.
  • Links (variety, number, type)
  • Delivery days/times (note: to test days of week, make sure to send all versions out in same week)

How it Works

Each version of one variable is sent to a small segment of your list or to your test group (if you have one). If you have a sufficiently large list, GetResponse can “split” the tests among equal portions of, say, 30% of your total list, but never more than 50%. After all, if only a few of your versions do well, you need to have someone left to prospect! The key is to always send each version of a variable to the same test group. If this is sounding like an algebra problem, let’s try one out.


Let’s say your list contains 1000 contacts and you’d like 30% of them to take part in the test. For this campaign, you only need to test two versions of your subject line, so:


  • A total of 300 subscribers will take part in your Split Test.
  • The first message will be sent to 150 subscribers, the second to the remaining 150.

Note: you are not allowed to send the first message to 100 and the second to 200 subscribers, hence the name Split Test (or Groups A & B).


  • The message with the best results will be sent to the 700 subscribers who did not take part in the Split Test. Again, you can choose which versions to send after reviewing the results, or let GetResponse do it for you automatically, based purely on the results.



Benefits of Split Testing

  • Improved response and conversion rates
  • Increased sales, revenue, and re-orders
  • Enhanced brand loyalty
  • Increased customer retention
  • Brand Reputation

And don’t forget, what you learn through Split Testing can be applied to future campaigns and newsletters, ensuring continuous improvement and professional development. You can also re-use successful titles, subject lines, messages, and graphics! That’s a HUGE time saver!


So please don’t waste time and money or risk your reputation simply because you think you have to hurry onto your next marketing task. Believe me, they’ll be waiting for you when you’re done! Split Test and go back to your work with the confidence that you’ve done everything in your power to ensure your success.


If you’re interested in learning more about how and why optimizing certain aspects of your emails through Split Testing may improve your results, please let us know. We have loads more information we can share! We’re also interested in your opinions, experiences, and general comments on Split Testing. Why do/don’t you use it? We need to know!

Become a marketing PRO!

Join more than 15,000 SMB marketers who get our top marketing articles straight into their inbox