|
E-mail Testing: You're Not Testing Enough
I recently led a discussion of a group of 35 Internet fundraising professionals. One of the common threads in the conversation was the desire to do more testing, but the inability to execute tests. I'd like to make some suggestions to help you, if you're in the same position. Testing is the essential element of measurement and improvement. Without a test, we don't know how to increase our response rate, pass-along rate, or average gift. Worse, we don't even know if this e-mail's results are better than the last because of something we did, or just a change in the mood of the country as a whole. In direct mail fundraising, you've probably heard some wise older person say, "We don't need to test that. It's worked for 25 years." While they may or may not be right in direct mail, they're as inexperienced as we are in Internet fundraising. The truth is, we don't know enough about what works and what doesn't, and there's so much to find out. If we miss the opportunity to test in one e-mail, we miss the opportunity to improve the next. OK, so I'm preaching to the choir. You already know how much you need testing. You wish you could test more. But there's never time. Here's a few ways to incorporate testing into your e-mail strategy with few pains: 1) Build a list of things you want to test. You should create this list without any particular e-mail campaign in mind. Just make a list of things like: html v. text, subject line, from address, etc. and tack the list on the wall over your desk, or in your conference room. 2) The next time you're discussing an e-mail campaign, look at the list and ask, "Which of these would be good to test with THIS e-mail?" If time is really tight, consider a simple test, like a variation of subject line. Save the complicated tests for those campaigns with more lead time. Know the limits of the technical support staff with respect to data segmentation and web page manipulation. 3) Before you submit the e-mail copy and landing page design for approval, draw up the test version and submit them together. It's not much harder than doing one, if you put both test and control on the same track. Select the data together, draw up the copy together, and design the two landing pages together (if needed.) Maybe you need separate thank-you letters - do them at the same time. 4) Execute the test with as much precision as you would for a direct mail campaign. Then, read the results, analyze them, and share them. Decide from that what the next logical step is. If you're too understaffed to plan testing into your e-mail strategy, contact NPA for help at 540-428-3640 or mailto:Rick@npadvisors.com. | |
|
|
|
