5 Tips for Split Testing an Email Campaign

Published by

Are you looking to enhance the performance of your email campaigns by dipping your toes into split testing?

If so, there are a number of important factors which should be taken into consideration before you begin. Therefore we decided to put together 5 top tips to prevent you from falling into some of the common traps associated with split testing:

Top Tips:

1) Change only one thing at a time

With a split test you are looking to find out why something works better than another version. Therefore if you change multiple elements how will you know what caused the improvement? Stick to one change each time.

2) Choose a reasonable sample size?

As a rule of thumb to be confident that your test results are valid you need at least 2,500 for each subject line being tested, and 8,000 for testing any elements in the email (Less people will see these so you need to send more). This would be the bare minimum to be reliable, and the more you can allocate to each test version the better.

3) Are your tests different enough for customers to notice?

There are loads of stories of how tiny tweaks to a web page make a huge difference to conversions. This isn’t the norm – for a split test to work there needs to be a difference significant enough for anyone to notice. Focus on big things like the from name, subject line, call-to-action buttons and images at the top of the email, not incidental changes to copy or anything far down the email.

4) Leave it long enough

You’d be surprised how long it can take before the results settle down for a test. Picking a winner after just an hour leaves you with a good chance of egg on your face when you look the next morning to see that another version has overtaken it! It is a drawback of split testing in email marketing as there often is enough time in your marketing calendar to properly test before you need to deliver, but if you don’t, you won’t get the growth you want. So how long do you leave it? We generally suggest at least 4 hours as a bare minimum, and longer if there isn’t a clear winner at that stage.

5) Repeat your test

For the ultimate confidence in your test, repeat it again. For an email campaign there are so many natural changes each time. A subject line test might have worked in one way last time, but can you consistently use that style on your next email? Did your test win simply because it was different to what customers are used to seeing?

For more information about split testing and how it can be used to maximise your campaign performance, get in touch today and speak to one of our experts.

About Nick Everitt

Senior Campaign Executive at Emailcenter, the UK's largest independent ESP.