Knowing When To Test

test

Lately we’ve delved a bit into methods of testing your email campaign and which aspects should be tested. Although we believe that most any aspect of your email send can and should be tested, the question that remains is when should you do it? The answer is: Constantly, but with limitations.

Throughout the course of the year, your open and click-through rates will be affected by a multitude of contributing factors, such as peak vacation days and holidays. Sometimes, though, you will notice a sudden surge or lull in these rates for no apparent reason. Is it an external factor? For example, did an ISP change their deliverability requirements? Could it be an internal factor, like a broken link or a poorly-worded subject line? When you begin seeing sudden changes in your rates, particularly for the worse, you need test what is broken and how you can fix it. Sometimes these changes can be for the better, in which case you can test to prove what you’re doing right and keep it up. Therefore, you need to constantly be testing every aspect of your email campaign to know where you stand and if you can improve.

With the belief that you should constantly be testing comes the downside of testing: using methods like the 50-50 A/B split test means you will need to slightly alter the aspect you are testing for half of your audience. If you begin testing the same aspect too often, the test will begin to lose its effectiveness. Your list will become accustomed to that aspect constantly changing, therefore becoming an alternate option on its own. For example, perhaps you are testing design variables within your email and your call-to-action button (CTA) is always blue. Then, you decide to test if making the CTA red will increase click-through rates. A week later, you want to see if green will have a stronger effect. Eventually, you’ve made this button every color of the rainbow and continue to see varied click-through rates. It could be that your audience likes one color significantly more than another, or the fact that the color is different every single time is eye-catching. Once you’ve begun to tread down this path, your results will become convoluted. With that being said, make sure you test one aspect at a time and do so every so often to ensure your results are exact and true.

Ultimately, knowing when to test boils down to whether you believe you’re getting the most from your email list. If you’re seeing inconsistency or significant room for improvement, then we highly recommend testing that aspect. Again, keep in mind to pick one aspect to test at a time. While you may want to test your subject line, layout, and content format, do not test all of these aspects in the same send.

 

Sarah Zibanejadrad
Inbound Marketing Coordinator, WhatCounts

[cta1]?