email testing

One of the many things that we do in the marketing aspect of our jobs is sampling. I am sure that better than 50% of annual giving offices test aspects in their direct mail program. Surprisingly, I find that many don't do similar in the phone program. Obviously it is much more complicated to test comparable scripts as you have a substantial additional variable in the caller(s) but it is an area where there is substantial opportunity for improvement and growth for most programs.

Even greater opportunity and less experimentation seems to exist in the email solicitation environment. Email is ideal for testing - it provides short term feedback with clear and direct results that can be tied back to the individual recipient. In addition, many email service providers (ESPs) provide tools that will allow for a/b testing wherein you can test a portion of the message, with subject line being the most common, and then send the majority of the list the more effective test variable. Thus, rather than mail or phone testing where we can make adjustments for the next solicitation, in email we can test and make adjustments within and between solicitations.

If you are not doing or taking advantage of this, start doing so today. Other easy portions to test are: time of day, day of week, sender (consider that many of us send solicitations from generic accounts but in our own actions limit the messages we open from non-people accounts), subject, format (text or html) and within format, the look - again how many friends or business partners send you email with a fancy layout and pictures included.

As a reminder, only test one item at a time and make sure that you send enough to make the results meaningful. Plan on doing one a month for the rest of the fiscal year. Create and layout a schedule for your electronic communications, identifying the audience the message and what you wish to test each month. Track and record your results on that same sheet and give yourself a resource to use in May or June (note that this is before, not after, the end of 2011-12) when you do the same thing for 2012-13.