- Do some analysis. "It's one of those old direct mail things that people forget about—doing a features, benefits and advantages analysis of whatever it is you're marketing via email," Jennings said. "Customers and prospects don't want to hear about just the features. They want more." She said it will be less difficult to plan a campaign—especially a drip campaign—if you can easily explain the benefits as well as how your product and service stack up against the competition's.
- Build a message map. Assuming marketers have already created several target audience segments, the next step is to take those features, benefits and advantages and decide which ones need to be targeted to which audiences. Jennings said this can help you plan an entire campaign based on those variables.
- Assess your re-engagement campaigns. Between 25% and 40% of all email lists are inactive, industry estimates find. Getting even a small portion of those subscribers to re-engage should be a goal for every marketer. However, most marketers go about doing so the wrong way, Jennings said. "Most say: "We noticed you haven't opened and clicked recently, so we're going to take you off the list if you don't respond to this message.' Not exactly a compelling message," she said. A better tactic is to remind people why they signed up for emails in the first place since many may have forgotten. "The key to every email relationship is the benefit the reader gets," Jennings said. To that end, see how and why a subscriber signed up and give them content that relates directly to their original action.
- Make sure testing results are statistically significant. The smallest sample size that works with A/B or multivariate testing is about 40,000 addresses, Jennings said—20,000 for the test and 20,000 for the control group. B-to-b marketers typically have much smaller databases than b-to-c marketers, so this may not work for some; but that doesn't mean they can't see valid results. It just means they need to do more sends to collect the same information, Jennings said. "Whenever you test, you just need to test for something that's repeatable," she said. "Test formulas—generic versus key topic subject lines, or TOCs in the preheader versus lower on the page or no TOC at all. Do several mailings and combine the data for a more statistically significant result."
E.J. Schultz on 08.28.2014