Do Figures Lie or Liars Figure?

Is it just me, or have you also read a plethora of statistics about email performance lately? Reputable sources report everything from open rates to how much email people want to receive. In fact, several of my columns have referenced such stats. How these stats should be interpreted and used to influence marketing decisions is a discussion worth having. As the number of sources quoting stats continues to expand, I can’t help but think some companies may make bad decisions. To quote an old adage: Do figures lie or do liars figure?

If Andy Rooney wrote this column, he’d say, “Let’s look at these email stats I have here on my desk.” They’re from a variety of sources, including: Quris, EmailLabs, and a recent NetCreations Webinar. My intent here is to put stats into perspective, lest we base bad decisions on them.

Factors that motivate signup for an email program: 41 percent of people surveyed said “sweepstakes or a chance to win.” (Quris)

When you see a stat like this, does it mean you should run out and add a sweepstakes program to your existing email campaign? Analysis: Be careful. Sweepstakes generally produce prospects interested in sweepstakes, not the underlying advertising offer. Consider presenting the underlying offer at the completion of the sweeps entry so it’s not dependant on entering the sweepstakes.

Factors that drive consumers away: 68 percent of people surveyed about what turns them off and causes them to want to unsubscribe said “emails [that” come too frequently.” (Quris)

Should you cut down the amount of email you send? If so, how much should you mail? Clearly, 68 percent is a big number. But too much mail is a function of relative content and whether your email is being viewed in the first place or is a victim of inbox clutter. At SendTec, it can take multiple mailings for a message to be seen. When in doubt, conduct a survey of your customer base to determine the preferred delivery frequency.

“In studies done by Forrester Research and NFO, email recipients preferred frequency of delivery is one email per week.” Thirty-one to thirty-five percent preferred once a week; 18 to 21 percent preferred once a month; and 8 to 18 percent preferred more frequently than once a week. (EmailLabs)

Though once a week was the most popular response, these numbers are all over the place. You really can’t draw a conclusion. As with the Quris finding, it depends on content and its timeliness. Someone receiving last-minute travel deals would most likely welcome a newsletter two or three times a week. A news junkie would welcome daily email. Other groups might prefer fewer communications.

Factors that drive email loyalty: 36 percent of people surveyed said “particularly interesting content.” (Quris)

Should you start developing additional content as a means to ensure people won’t unsubscribe? I’m a believer in content, but you should first examine whether you’re using the right content where it’s appropriate. Don’t summarily discard your plan and start over. It’s no secret that the more relevant the content is to the recipient, the higher the open rate and CTR.

In a recent NetCreations Webinar, 24/7 Canada revealed its average CTRs were 3 percent for English-speaking recipients and “much higher” for French Canada.

Should you run out and rent Canadian email lists? Should you translate your offer into French and test French Canadian lists? In the email environment, it’s so easy to find a list and start emailing Canadians. But there are lots of issues to think through before sending your first marketing message. Can you legally sell in Canada? What do you know about the exchange rate? Do you need a fulfillment operation in Canada, or can you ship from the States? What laws govern sales made via Internet and phone? What are the shipping rates to Canada?

Now that we’ve looked at how stats can be analyzed, some suggestions regarding what to do with them.

It’s easy to be mesmerized by stats and start questioning your decisions. It’s easy to read a statistic and say, “If 70 percent of marketers are doing that, I should, too.” Of course, it’s not that simple.

Every stat should be reviewed in the context of what’s relevant to the recipient of your email and your business model. You wouldn’t send jokes to doctors or baby-related tips to CEOs, would you? Stats are dangerous unless they’re in context.

Ask yourself the following questions to determine if a particular statistic should be considered in terms of making changes to your email marketing program.:

Who’s the audience surveyed, and how was the data derived? If data came from Webmasters or IT people and you sell a consumer product, the stat may not be relevant.
• What’s the source? Is it a reputable research company or an “esoteric review of the literature,” which may or may not be valid or even current? This space changes rapidly.
• Is the stat a big enough number to make sense? If a survey indicates only 10 percent of people surveyed appreciate newsletters, that 10 percent figure would be a lot less important to me than if 80 percent of them appreciated newsletters. Most of the time, I pay attention to stats only when 50 percent or more of those surveyed respond.
• Does the stat apply to your business? If you sell business-to-business (B2B), do business-to-consumer (B2C) stats apply? Probably not. Even within a B2B survey, some stats may apply and others may not. Determine if any given statistic applies not only to your industry but also to your business.
• If I want to act on the statistic, what would I do? What changes would I make that would lead to improved sales, revenue, or profits or decreased expenses? Think through all the ramifications of changes you’re considering.

Next time you get a newsletter loaded with the latest stats, find the ones likely to have the most impact on your business. Then analyze them.

Keep reading…

Want more email marketing information? ClickZ E-Mail Strategies is an archive of all our email columns, organized by topic.