|
Many marketers can miss the big picture of email marketing. While email is cheap to send and quick to answer, the biggest and best advantages of email are in the ability to know exactly what works and what doesn't work. Email is inexpensive to send and offers you so much more information than direct mail in understanding your audience. Your turnaround time is lessened from weeks to days when running tests.
Yet in the world of email marketing, very few marketers take the time involved to test drive their emails first. Many will ask if it's worth the effort. If you can test an email that tells you that putting the order button at the top of the email garnered 50% more click throughs to the website than the email with the order button at the bottom of the email, isn't that worth looking at? Whether you are spending hundreds or millions, can you really afford to just throw money away?
The only way to truly know what works with your audience it to test them. And I don't mean giving them Math quizzes. Test out their preferences, their emotional triggers - do they respond to negative or positive subject lines, do they like HTML or text, long or short formats. You can learn valuable information about your audience by testing variables in your emails. Here are some helpful hints to guide you in your testing forays.
1. Start with just the basics.
Testing your email is always a good thing to. Sometimes, the best thing to do is test the simplest portions of your emails, such as the formatting. Many emails are never looked at simply due to faulty HTML. Testing your email in a variety of email applications, from Outlook to Yahoo to AOL, can make a drastic difference in your overall response rate. It's too easy to just assume that your email looks the way you want it to. How will you know that unless you test it first?
2. Keep it under control.
Try not to go buck wild and test everything in your email all at once. If you send two completely different emails to different groups, you won't have a clue as to what worked and what didn't work. Choose a single element to test, one at a time (such as long versus short format) and see how your audience responds to that one variable before applying it and then testing a new variable. Keep your tests controlled. The only way to judge how well one thing works is if you have something to judge it against.
For instance, let's say you have a new product and are sending an email to your customer list to tell them of the new product. The e-mail's main goal is to get them to purchase the product. To help further that goal, you are offering a discount and you want to know what the best way is to offer that discount. Will they respond better to a discount that offers 30% off or a discount that offers $20.00 off? If this is your variable, it should be the only thing different in your email.
3. Timing is everything.
It's important when sending out two (or more) test emails, that the emails go out at the same time, the same day, the same season, planet alignment…you get the picture. Sending Email A at 8am and Email B at 5pm will give you back very different results, and that also falls into the category of more than one variable at a time. (For a quick hint, many researchers say that Wednesday is the best day of the week for responses.)
4. Know the response you need.
Statistics show that 30 to 50 responses will give you a fairly accurate count on what works. 100 responses will give you an almost 95% accuracy count. So you can either just split your list in half (randomly of course) or you can separate a smaller list to test out the elements on first and see which element does better, before sending to the larger list.
How big of a list do you need? That depends on your click through averages. Let's say your list is 10,000 emails. Your average click through rate is 8%. Consider how many responses you need (we'll say 100). With an 8% click-through rate, 1300 emails sent would result in 104 responses. That would feasibly give you a very accurate count for your test. If your click through rate is higher, the number of emails you need to send to test would be lower. And obviously, the larger of a response you get, the better your results.
5. Don't just accept the grade. Learn from your tests.
Generally speaking, when we took tests in high school, we were more concerned with the grade we got rather than what we learned. I mean, really, how often did you take your wrong answers and go searching for the right ones? However, when it comes to testing your email marketing efforts, your goal for testing an element has to be to actually take the lesson learned and apply it. Otherwise, you're testing just to test and who really wants to do that? You have to be prepared to apply your newfound knowledge going forward.
|
