<img height="1" width="1" alt="" style="display:none" src="https://www.facebook.com/tr?id=799187456795375&amp;ev=PixelInitialized">

Scientific Email Marketing: 6 Steps to Improve A/B Testing

January 7, 2015
By Tom Burke

scientific_email_marketing_ab_testingImagine if we substituted proven methods for gut instincts. If you were a scientist trying to figure out which hand soap is also the best germ killer, you wouldn’t ask five people to wash their hands with a different soap and make a decision based on whose hands looked cleanest. Rather, for a test to have integrity, you need to employ reliable testing techniques. Email marketing is no different.

By accounting for variables, controlling the testing environment, and accumulating a significant data sample, employing uniform A/B testing methods will help you optimize your email marketing ROI.

Here are the top six steps we recommend:

Step 1: Carefully define your question.

This sounds deceptively basic, but narrowing the focus of your test is essential. Big, over-reaching questions are difficult to test. Instead, try to winnow your test down to a very specific this-or-that question, such as, “Which will convert better: ‘Become an Official Hand Soap Tester’ or ‘Help Keep Your Family Healthy’?”

Step 2: Gather information.

Once you know what you’re testing, conduct some preliminary market research. To use our soap example, ask yourself what you know about people who use soap? What data can you build on? What data do you still need to uncover? This data will highlight the variables you should be testing.

Step 3: Form a hypothesis.

As you might remember from your high school science courses, a hypothesis rephrases your initial question into a prediction for your test. State the hypothesis in clear, testable terms to isolate the variables that may impact your test. For example: “In promotional emails sent to mothers of young children, the call-to-action ‘Help Keep Your Family Healthy’ will convert better than ‘Become an Official Hand Soap Tester.’”

Step 4: Test your hypothesis.

Keep a tight focus on the variables you’ve identified. If you’re testing the call-to-action, don’t compromise your data by introducing other variables, such as new headlines, new graphics or new content, until you’ve collected a significant data sample on the call-to-action variable. In testing, the more email data you accumulate, the better.

Step 5: Analyze your data.

Record your findings as precisely as possible; even small trends can reveal huge opportunities. Perhaps your data shows that emails with the first call-to-action had more shares, but you received more click-throughs with the second call-to-action.

Step 6: Draw conclusions.

Was your hypothesis correct? Depending on your data, new questions might arise (i.e. Why was the first call-to-action shared so much more?), which you can try to answer with a new experiment. You might also find an opportunity to test a new variable, such as “Should the word ‘healthy’ be used in the subject line?”

While it may take some time and effort, A/B testing pays off. This simple practice can help you discover information about your contacts that helps you create more effective strategies for the future.

Ready to modernize your strategy and increase clicks? Download this guide to learn how to acquire and use customer data using TowerData’s Email Intelligence.

TowerData's Email Intelligence Services

Share Your Comments