Test for Success

It always surprises me that as marketers we give lip service to the importance of testing, but many of us don’t take the time to fine-tune our e-mail programs. Just think, if you are able to identify a few elements through testing that resonate with your customers and then employ those insights in your ongoing program, the improvements may be sustainable over time. And, studies show that marketers who test have higher ROI than those who don’t.

Approaches to Testing
If your program generally works well and compares favorably to industry metrics, you may want to isolate individual variables and conduct some A/B tests to determine which is more effective. A/B testing is easy to set up and doesn’t require advanced analytics to determine success.

If your program is not meeting your objectives and results are below industry benchmarks, it may be time to take more drastic measures. You may want to try a major overhaul and test multiple elements at the same time. This is similar to direct mail where you might develop and test an entirely new format and presentation against your control.

In either case, you want to structure your test so you have reliable results. If your list is sizeable, it is relatively easy to take 5 percent to 15 percent of your file to conduct your tests. If your file is small, it’s more problematic. Structure your tests so you’ll have at least 100 responses to measure for each test group. Depending on your goal, responses might be opens, clickthroughs, clicks-to-opens or conversions on your site.

If your testing is straightforward, you also can pretest prior to a campaign and then roll out to the balance of your list with the winner.

Easy Tests to Consider
Let’s start with some ideas that are easy to test. While they are simple, they can have a significant impact on results.

• Day of week or time of day. You likely send your e-mails on a consistent time schedule in terms of day of week and time of day. Most marketers concentrate on Tuesdays through Thursdays. If you follow this schedule, perhaps your recipients are inundated with other e-mails. Will you stand out if you change your sending schedule? Examine when the most opens and clicks occur, and determine whether that gives you additional insights on when to test.
• Subject lines. There are some interesting and conflicting studies on subject lines. Some find shorter is better, and others find that longer subject lines yield better click-to-open ratios. Test this for yourself. Or test different approaches such as highlighting benefits, including personalization (name, segment or geography) or asking a question.
• The call to action. This includes test buttons vs. text links, the color of the button, the wording of the text link, the number of times you repeat the call to action, or the placement of action item. While you might expect that buttons would perform better than a text link, your recipients may be reading your messages with images turned off.
• Discount offers. See whether a percentage discount or dollars off a purchase makes a difference.
• Snippets. Does a marketing message that supports the main theme of your message included at the top of the preview pane in addition to administrative messaging (such as “click here to view”) increase open rates?
• Frequency. If you currently e-mail more than once a week, create a group that receives only weekly e-mails, but measure these results over a longer period of time.

Testing That Requires More Planning
These ideas require more marketing, creative and design involvement, but can make a difference.

• Placement of the second column. Many e-mail marketers use a two-column design. The smaller column could include supplemental information, product features, banner ads or other information about what’s happening on the site. Understand whether a smaller left or right column works for you.
• Inclusion of main site navigation. Do results improve if you add major departments, a search box or inclusion of your 800 number? If you already include this information, conduct a test omitting this information.
• Landing pages. Experiment with the content and design of your landing pages. Hint: Often the simplest designs work best.
• Long vs. short copy. Whether you have a promotional or informational e-mail program, you can test whether your audience prefers longer descriptions and articles, or more concise copy. Obviously, you need to provide enough content to whet the appetite to click through, but some people want to be romanced or provided a fuller picture.
• One image vs. multiple images. Here you want to try the complete opposite of what you regularly do. If you usually have one strong featured image, try multiple images—or the reverse.
• Segmentation. Reports consistently show that marketers who use some segmentation see a higher return on investment. If you haven’t done this in the past, you’re missing out on a big opportunity. Try it! And this doesn’t have to be overwhelming. Start with just a few segments based on what you know about your list, and tailor your offers or content based on this information. You will need to measure this over time.
• E-mail to third-party lists. There are slow and steady improvements in lists available for acquisition and lead generation. Even if you have tried this in the past, now may be the time to test the waters again.

You’ve heard the adage, “Nothing ventured, nothing gained.” Testing to understand what works best with your audience yields results. Let me leave you with one example that I hope will inspire you. One marketer with an e-mail list of 3 million needed to test an important offer. It selected 10 percent of its list and tested 30 different variables to find the winning subject line, offer and design configuration. The winning combination worked extremely well.
You don’t have to be as complex in your planning, but make a resolution to test.

Regina Brady is president of Reggie Brady Marketing Solutions, a direct and e-mail marketing consultancy. She can be reached at (203) 838-8138 or reggie@reggiebrady.com.