Among the many strengths of e-mail is that it enables marketers to track and report on subscribers’ behaviors. This also makes e-mail an ideal medium for testing and optimization. Lists can be segmented and tracked separately, A/B tests can be run on small groups, and winning tests can be rolled out to larger audiences.
Savvy marketers understand the importance not only of testing to increase performance but of choosing the right types of tests to run. They also understand that what worked previously or for one e-mail program may not work today or for another program. This is the reason that the answer to most questions regarding e-mail marketing is “It depends; test it.”
The secret to successful e-mail testing is to develop a hypothesis of what you want to test and to keep a log of the results to see if your original assumptions were correct. Change only one thing per cell, and be sure to track all the way through to conversion.
One of the key challenges with optimization testing in email is finding the right element to test, one that will yield not only significant results but also repeatable and actionable results. Nothing can be worse than doing an A/B subject line test and rolling the winning subject line out to your full audience, only to discover that you can’t determine what about the subject line drove increased responses, so that you can’t repeat it.
Below are a few areas where simple tests can lead to significant results:
* audience segmentation. While it is tempting to try to apply what works among one group of subscribers to every person on your list, different segments often respond to different stimuli. Before making any changes to subject lines, calls to action, or offers, create segment groups of similar subscriber behaviors. These can be by acquisition source (quick sign-up vs. purchasers), length of subscription or activity, or preference data. Apply these segment groups when you do your A/B split runs and you’ll quickly have a matrix that will show how the results compare against the different types of subscribers on your list.
* subject lines. Subject lines are the primary driver of whether subscribers will open a message, and the expectation is the more readers, the higher the conversion should be. There are essentially three main elements you should test in subject lines:
· * brand-specific subject lines (company or product name)
· * action-oriented subject lines (for instance, “limited time only” vs. “shop now”)
· * benefit-driven subject lines (offers and promotions such as free shipping).
Add to this personalization and placement (front, middle, and end of the subject line), and you end up with an interesting testing matrix. Prominence and placement are controllable variables and are therefore ones you can use to apply to future messaging.
* calls to action. A strong call to action has three key elements: the action you want the reader to take; the words you use to issue the call; and its physical appearance (font, image, location). As a command, “click here” is beautiful in its simplicity, but it falls far short as a call to action because it doesn’t tell your e-mail readers what you really want them to do. With the recent recession, many marketers are testing their call-to-action messaging (“buy now” vs. “save more”) to reflect the current concerns of subscribers and price sensitivity.
Even simple A/B can yield powerful results, but you need to document the findings so that they don’t get lost over time. Testing should not be ad hoc, but rather a process of learning what drives your audience to respond at higher rates so that you can improve future messaging. Be careful not to test for the sake of testing but to have an idea of the expected outcome, and test to validate your assumptions. Then apply what you have learned to future campaigns, and you will truly have optimized your results.
Stefan Pollard is senior strategic consultant for marketing solutions provider Responsys.
