One of marketers’ common downfalls with email marketing is not A/B testing. Or, not A/B testing effectively.

The idea of A/B testing is to work out which variation of your email works most effectively. Then you can determine the best-performing version to send to the majority of your database. We use samples within your database to test the different elements.

It sounds simple, but there are some common mistakes that can have a negative effect on your email results and your ROI. Below, you will see 9 best-practise tips followed by three examples of A/B testing that could help your business.

9 Tips

1.    Test the important elements for YOUR business
It’s easy to fall into your own trap where you try to test everything, and you might find yourself testing things that don’t really affect your business. Identify the most important elements within your email and only test them.

Scenario
A kitchen company has a strategy to generate leads through phone calls or email enquiries from the email campaign.

The measures of success for this email are:
•    Open rate
•    Phone calls
•    Email replies

There is no need to start testing the social media buttons to see how many recipients click through to the social media* pages.

*Some kitchen companies may have a strategy around driving traffic to their social media page, but in this scenario, it holds little value for the business.

2.    Test one element at a time
Businesses may have three of four different elements of the email that they want to test, but you should only test one element at a time. Otherwise, the results become subjective and the conclusion lacks validity.

You want to know for sure which version of each element is most effective, so if you are testing the call-to-action button, ensure the subject line and content is identical.

3.    Standardise the time of the day
When it comes to email marketing, the time of the day has a massive impact on results. For that reason, you must send all versions of the email at the same time. If one email goes out 30 minutes after another one, that could skew your results considerably.

4.    Sample size
Industry best-practice is to test each variation within a samples size of at least 1,000. Any smaller and the numbers are too small to produce an accurate conclusion.
Depending on the element you are looking to test, this may require sending to sample sizes more than 1,000 contacts.

Example
A beauty salon wants to test the positioning of their call-to-action button within the email. This means at least 1,000 contacts need to see the call-to-action button, which means that over 1,000 contacts are required to open the email.

If the beauty salon has an average open rate of 20% (1 in 5) they will need to send each variation of the email to a sample list of 5,000 contacts, based on the assumption that 20% open the email and 1,000 contacts view the call-to-action button.

5.    Split the list evenly
When A/B testing, it’s important you split the lists evenly. Most marketing tools offer an A/B testing option today that randomly splits the list evenly, but you must ensure one email is going to active contacts and the other going to inactive contacts. Obviously, this would skew the results, too.

6.    Consolidate the results
It’s easy to do an A/B test one month and then not document those findings, so they are not easily accessible next month, next quarter or next year. We suggest you
create a simple report that clearly outlines the following:
•    Elements tested
•    Variations tested
•    Results
•    Winning variation

7.    Test the subject line 
The most important A/B test to improve your open rate is to test the subject line – you can read more about writing an effective subject line and pre-header text here.

We recommend you test personalising the subject line.

Example
•    Email 1: Here are our 9 top email marketing tips
•    Email 2: David, here are our 9 top email marketing tips

8.    Test your Call-To-Action (CTA)
The call-to-action is commonly the most valuable element of your email as it generates the positive engagement and response.

It’s important you test the best position for the CTA button, taking into account your open rate on your mobile.

You may find that the CTA appearing above the fold (on view when the email is opened) has very different results to a CTA appearing below the fold (requires the contact to scroll down through the email).

Example
•    Email 1: Place the CTA towards the top of the email. This is so the CTA gets the most prominence within the email.
•    Email 2: Place the CTA towards the bottom of the email. This is so the contact sees the CTA after engaging with the content, which could provide some substance to the product.

9.    Test testimonials
Testimonials are designed to generate trust between the businesses and consumer, validating the service being offered. Run an A/B test on the inclusion of a testimonial within your email as if could improve the conversion rate.

Final hint
Remember, small details matter when it comes to email marketing. You may have concluded the CTA works most effectively when it appears above the fold, but you can also test the colour of the CTA to see if that improves it further. It may sound like a needless test, but it can have a big impact on your bottom line.