Walking into an ice cream shop presents you with many possibilities, as there are numerous flavors and almost limitless combinations available to you. To make the decision process easier, you have the ability to taste samples of a few flavors before making a purchase. From the ice cream parlor’s point of view, whether you go with vanilla, rocky road, mint chocolate chip, or strawberry, there is no specific flavor that would yield a less than perfect conversion at the register. From your point of view, the tasting process has a high impact on the type of purchase you will make.

Driving conversions over the Internet requires a similar philosophy to ice cream sales, only taken one step further.

Email testing should be as routine as sampling flavors at your local ice cream parlor. E-mail marketing, similar to ice cream, lends itself to testing, because you can discover interesting results from sampling among a large pool of variables. If your email subscriber lists are large enough, I recommend sending a small portion of tests and control versions of your e-mail to 20% of your list, then deploying the remaining 80% a couple days later to the winner of the test (or even better, try a random 10-10-80 split).

Testing garners proven results, according to Jeanne Jennings from ClickZ:

“Not every test will result in a lift in response. But even if you can improve performance only 1 or 2 percent with each test, you’ll see your results gradually increase over time. Continually testing increases your odds of hitting the jackpot with a lift of 20 percent, 50 percent, or more.”

Planning for repeated tests will enable you to explore the full potential of email testing. To help you get started, I’ve compiled a quick list of email marketing variables that can be easily tested:

• ‘From’ address
• Subject Lines
• Calls to action and their associated graphics
• Copy (long vs. short)
• Creative design
• Day of the week
• Time of day
• Frequency of mailings
• Offer

With testing, it’s all about measuring customer value over a period of time. Planning for repeated tests will enable you to determine which groups of customers prefer specific offers, subject lines or design layouts – and using that information will help you realize the full potential of your email program goals.

Tags: 10-10-80, ice cream, split, testing

Trackback URI | Comments RSS