Substance and Style

Email marketing is one of the most effective ways to promote your business. It has one of the highest Returns on Investment of any marketing method, with the latest average being over $36 earned for every $1 spent. To get that high ROI, you need to optimize your methods. A/B testing allows you to figure out quickly what’s working and what isn’t. By sending two versions of your email to two different segments of your audience, you can see which one performs better and avoid spending time and resources on less effective campaigns. The basic idea is to compare open rates, click-through rates, and any other conversions you’ve established, and then abandon whatever techniques don’t help your bottom line. Here’s how to set it up:

1. Define your goal. Look at the big picture of what you’re trying to achieve with your email marketing in general (increasing sales, driving website traffic, boosting brand awareness, etc.) and consider what is getting in your way (low click-through rates, low open percentages, etc.). Focus on improving whatever is holding you back from your overall goals. For example, you could optimize your subject lines to get more opens.

2. Figure out what you should test. Only pick one element to test at a time. Otherwise, you won’t know what’s responsible for any changes in results. Common things to test include: subject lines, pre-header text, Calls To Action, send day and time, personalization, and design elements within the body of the email. Pick one thing that could be responsible for the metric you’re trying to change. 

3. Create two versions of your email, with the only difference being the thing you’re testing. 

4. Create two audience segments. Do not use existing segments you’ve created that are based on differences in behavior or demographics, as this will lead to confusion about what is driving the differences in results. Instead, create equal lists with equal numbers of representatives from those segments. Try to make the lists as similar as possible in terms of behavior, past engagement, and demographics. Make sure you’re using a large enough sample size to get meaningful results. 

5. Send the emails. Make sure you wait to analyze your results until you can be reasonably sure you’ve gotten most of them in. The length of time you should wait depends on your own list, how frequently you send emails, and how long it usually takes your audience to open them. Obviously, you shouldn’t send your email at 2 AM on Sunday morning and conduct your analysis at 8 AM that same day. But, if you usually send one email a week, you also shouldn’t wait a month to analyze your results. There is some guesswork involved and it may take some trial and error, but you should be able to make a pretty good guess about when to check your results based on your usual email performances. 

6. Measure the results and learn from them. Analyze the differences in opens, click-through rates, sales, or whatever else you’re trying to measure. Figure out if your changes account for the differences in the two groups. 

7. Tweak future campaigns to use what you’ve learned and optimize your performance.

8. Repeat the process on a regular basis  to continually improve your results. Make sure you’re keeping track so you can check overall patterns and make sure you’re headed in the right direction.