If you’re sending e-mail campaigns, then you really must make sure you are testing their effectiveness regularly and making tweaks accordingly. Doing so will lead to more clicks, opens, and sales. Here’s how I recommend you approach testing your campaigns for better results.
- Develop a control e-mail and test it. Create an initial e-mail and note the results it gets. Then, decide the one thing you want to test. It could be the offer or the subject line. Choose just one. Send the e-mail again and see what results you get. When you have an e-mail that gets better results than your control one, it becomes the new control e-mail that you’ll do further tweaks to. I do this process for my clients like TCBY, Pie Five Pizza, and Hurricane Grill and Wings to name just a few. This is a really powerful tool that enables you to get great results.
- Test your mailing list. For lists over 1,000 I like to split it into two groups. Each group gets a different e-mail message. For much bigger lists (10,000 and up), I like to send a test message to a small portion of the list, say 20 percent. If the message pulls well, then I roll it out to the remaining 80 percent.
- Test different variables. Here’s where you can get really creative. Test different offers, different design styles, shorter or longer message, personalization, subject lines, etc.
- Test your mailing days. The time and day your e-mail is sent can effect your response rate. Consider your target market and what day and time makes the most sense, then test it.
- Test your frequency. This refers to how often you are e-mailing. Keep track of when you are sending your e-mails as well as the responses, bounces, and unsubscribes. Look for trends that will help you figure out the right amount of e-mails to send to your list. If you e-mail too infrequently, you’ll likely loose your recipients’ interest. If you e-mail too frequently, chances are they’ll opt out. Find a happy medium.
- Keep testing. This is an ongoing endeavor. To keep a campaign successfully, you have to keep testing. Remember though to only change (or test) one thing at a time, that way you know for certain the results your change produced.