1. Choosing your "control"
2. Deciding how many test to run

1. Choosing your "control"

In order to get accurate results when you're testing the effectiveness of your promotions (or anything else, for that matter), you must test only one element at a time. If you try testing several different elements at the same time, you'll have no way to determine which change improved your response.

For example, let's suppose you decide to test two different e-mail promotions. The first is structured as a long, formal letter to your customers written in HTML, with the subject line, "Dale could you give me your opinion?" The second e-mail is a short, punchy note written in a personal tone and delivered in plain text, with the subject line, "Dale, could you help me out?"

After testing both letters from a small list of your customers, you discover that the second test is bringing in a better results, so you decide to launch that letter to your entire list. The problem is that you haven't learned WHY the second letter was more successful. Was it the subject line? The text format?

This is why it is so important to establish a "control" letter, and to test different versions of your control letter with only one thing changed at a time.

2. Deciding how many test to run.

You can run as many test as you feel are necessary to tweak and perfect your email promotion. Recommendation--run a minimum of three tests before launching e-mail promotions.

In general, you will want to test at least two different letters and two different subject lines for any e-mail promotion you plan on sending. The testing setup would look something like this:

By running these three test, you will be able to get a good feel for which elements of your e-mail promotions are generating the best results so that you can launch the most profitable combination of letter and subject line. The more testing you do, the more targeted and effective your promotion will be.

However, if your opt-in list contains less than a few thousand subscribers, you may need to watch trends over a period of a few months before you'll have a solid feel for what your subscribers prefer in terms of promotion styles, format, and voice.