By Nerea Boada, on 1 March 2017
If you want to improve your conversions in 2017, prepare to become best friends with A/B testing. By being systematic when testing different variables in your campaigns, you will increasingly optimize your results and, in turn, improve your conversion and return on investment. In this article, we will look at what exactly an A/B email marketing test is, what not to do, and how to use them to optimize your campaigns.
What is an A/B test in email marketing?
Running an A/B test in an email marketing campaign is based on a very simple process: sending out two version of the same email, which differ in one single variable, in order to see which of the two people are more inclined to open and click on.
There are several variables that can be tested, such as the subject, the body text, the Call To Action copy, the color of the CTA button, the images, the offers or even when you send the email. What you are looking to do is make small, identifiable changes that improve the result you are focusing on in each situation: the number of people who open your email, the number of clicks on a certain link or the conversion to sales, for example.
What NOT to do in your A/B testing
Know that we know exactly what an A/B test is, let’s take a look at the mistakes we must avoid for it to work:
- Changing more than one variable at a time. This is one of the worst mistakes you can make. If you want your tests to work, do not rush yourself and change one single variable in each test. The reason is very simple: if you change the email subject, the body and the color of the CTA button at the same time, for example, you will have no idea which one of the three it was that improved your results. Although in the end you will end up testing all possible variations, be patient and do them one by one.
- Using email groups that are too different. These A/B tests are based one sending the emails to two different groups of users to be able to compare the results later. If the two groups you send the email two are different from each other, you will not be able to compare the results nor reach any conclusions on what works best. To avoid this problem, be sure that your two groups of users are as similar as possible.
- Using email groups that are too small. If you do not run your test on a large enough amount of users, something as simple as bad luck can completely distort your results. Just like in scientific studies, you need to have a large enough “sample group” for your data to be reliable.
- Not taking risks. These A/B tests are precisely that - your testing field. Leave your comfort zone and try out all kinds of ideas. Those who never take a risk will never win!
How to run an A/B test in an email marketing campaign
1) Define your goals
Just like in any general email marketing campaign, you need to know beforehand exactly how you are going to measure the results of your A/B tests. What exactly are you trying to get out of it? Improve your opening ratio, increase your sales via email, get users to click on a specific link…?
2) Decide who your test subjects will be
This step is crucial to guarantee the success of your test. As we have already seen, you need to have a large enough group of people to be able to reach and conclusions, and they need to be similar to each other. On the other hand, however, the goal of your test is to know which version works best, so it does not make sense to use your whole database.
To successfully run an A/B test with the best possible results, choose a percentage of your database (like 20 or 30% for example), and send one version of your email to one half and another version to the other half.
It goes without saying that for your tests to work you will need to have a large, high quality database of user emails.
3) Put your test results into action
Got your results? Great! Now, all you need to do is send the winning version to the rest of your database. Be sure to keep note of your conclusion to use in your future email marketing campaigns.