Useful Analytics
If you have an analytics person on your staff, you should always be looking for way for her to improve your marketing operations. For example, if you send promotional e-mails to registered customers, she should take advantage of the wilderness of numbers that can be analyzed with profitable results. The trick is to figure out which numbers to analyze, and what to do with the results.
For example, a retailer sent e-mails 70 times a year to an average of about half a million registered subscribers. The total annual online sales from these emails were $21 million, which represented 20% of the sales from their retail stores. So what is wrong with that? Nothing, but they decided to do some analytics anyway.
A couple of interesting numbers popped up. During the year, 102,683 subscribers elected to unsubscribe — to stop the emails coming altogether. These unsubscribers were balanced by an almost equal number of new subscribers, so the total audience for the emails remained about the same. In addition, 1,010,092 of the emails sent to subscribers during the year bounced back and could not be delivered.
The question arose, “What were these people doing before they unsubscribed or their email addresses become undeliverable?” The answers were not surprising. Most had bought nothing at all before they disappeared.
Some, however, had been very active buyers. Altogether 9,712 of them had made purchases before they left – and 2,160 had bought two or more times. Those in the group that left, collectively, spent almost $1.8 million in the year before they disappeared. Clearly, these folks did not leave because they did not like the retailer’s products. They left for some other reason. What was it?
That brings up the second half of the analytics problem: What to do with the results? The fact is that no one really knew. The retailer had no program in place to ask unsubscribers why they left. The answers might have been interesting. So the first result of the analytics was a recommendation: Set up a program to create a Web survey question for all active buyers who ask to unsubscribe. The answers could be very important – and might lead to keeping some of these valuable people as customers, rather than having them disappear down a memory hole.
The analytics yielded another valuable piece of information. About 9% of the half million subscribers bought something during the year. Ninety-one percent bought nothing at all. How did the emails sent to the buyers differ from those sent to the non-buyers? Answer: there was no difference. But there could have been.
All the information that the retailer had on most of his subscribers was their e-mail address. He did not even have their names. For this reason, he could not personalize the e-mails (“Welcome back Arthur…if you’re not Arthur click here”). But for the buyers, the retailer had the names and addresses (how else did they pay for the products and have them delivered?) Recommendation two from the analytics team: Send buyers different emails than those sent to the non-buyers, and personalize them.
The analysis of the purchases made by the unsubscribers led to another analytical question; How many buyers do we have who have bought multiple times during the year? That answer was the most interesting of all. While most of the buyers bought only once during the year, there were 38,875 buyers who bought twice or more. They spent a total of $17,382,096. The average per person was $447 each compared to about $100 from the one-time buyers.
What was done to reward or encourage these multi-buyers to make sure that they did not disappear? The answer, of course, was nothing, since before the analytics was done no one was aware of the existence or identity of these valuable people. The analytics group’s recommendation was to find a way to recognize these buyers so they feel appreciated, and will keep on buying.
How long did it take to do this analysis? About three days. It seems so obvious that one might say, “Why wasn’t it done before?” Good question, but the answer is a simple one: all previous analysis was done on the e-mail campaigns (opens, clicks, conversions, deliverables, etc.) and none at all on the subscribers.
So if you have a good analytics person who is available to work on your email marketing program, make sure that you allocate some of her time to analyzing subscribers and buyers – and even more time to the question of what to do with the results of her analysis. If you don’t have an analyst, you might think about outsourcing this kind of work occasionally. The ROI should be substantial.
Arthur Middleton Hughes is VP/solutions architect at KnowledgeBase Marketing. Arthur is the author of Strategic Database Marketing 3rd Ed. (McGraw-Hill 2006). He may be reached at [email protected].
