Marketing campaigns benefit from email split testing, as good tests provide insight to improve future campaigns. When it comes to split tests for email marketing, we have a few tips that will help you get the most out of it:

Decide What Type of Result You Are Testing For

Open rates, click rates, and end results are all stats that you can gather and learn about with split testing. End result refers to the action that you’re looking for on your site: it could be a purchase, a survey response, an event registration, or something else. Trying to test more than one of the stats will muddy the findings, so choose only one and measure the results.

Testing Open Rates

This type of test evaluates what leads more people on your list to open an email. Nothing in the email body factors into this test, as your contact sees the body of the email only after they’ve already opened it and are counted in the test results. Factors you can modify in an open rate test include: subject line, preview text, time of day that the email is sent, day of the week that the email is sent, and who the email is from.

Testing Click Rates

Click through tests look at the contacts who opened your email, how many of them clicked on content, and what they clicked on. None of the factors or variables used for an open rate should be split tested when you evaluate click rates, so keep everything consistent up to the open. Variables to test for click rates may include: email template design, the tone of voice (is it humorous or serious?), animation, button styles, graphics, images, products, content, or copy length.

Testing End Response

This test is looking at the action taken as a result of the email. If I send a sales promotion to List A and List B, I want to know which list produced more sales. The variable you test here could be a split test of the page you link to from the email. For example, you may want to test the difference between linking your call to action buttons to a product page or a category page. Alternatively, you may link both parts of the split test to a registration form, but change the format of the form.

There are times when a test for end response involves changing one of the variables normally used for open rate or click rate tests. Why would you change one of those variables and measure end response instead of open or click rates? In this case, you’re looking for the continuity of the experience leading from the inbox to the action you want your contact to take.

An intriguing subject line that generates more opens would win an open rate split test. However, it may resonate with people for a reason unrelated to your end result. For example, the subject line “Open this email for a surprise” could generate more opens than “We’re running a special sale on dishwashers.” The dishwasher subject line may generate less opens, but those who do open will be more relevant and interested buyers for dishwashers, thus leading to more results.

What to test depends on what you want to learn about your email program and your contacts. Test often, test wisely, and test with a strategy in mind.