This is the second post based on our recent conversation with Neal Goff, a longtime publishing executive with extensive experience with direct response who is now consulting to publishers through his firm, Egremont Associates. Because we believe that trade publishers have to become B2C marketers like they have never been before, we think it is important to take on board the fundamentals of managing contact with consumers. Neal is explaining them to us, and we’re trying to pass that understanding along.
In the first post, we wrote about the importance of name gathering, the data that should be captured (recency, frequency, monetization, and affinity) and the challenges publishers might expect to face putting their marketing database together. In this one we’re going to say more about what seems to be Neal’s favorite subject: testing.
By the simple device of changing one element of an emailing effort, one can learn which choice of that element produces a better result. But you must change one element, not more than one. If you change more than one, you wouldn’t know to which element to attribute any noticeable change.
There are three major components of any mailing that can be tested: the list, the “creative”, and the “offer.”
The list is the names you choose to mail to, which is comprised of the group you selected from (recent purchasers, or recent purchasers of a particular book, or people who said “send me material about this author”) and the selection criteria you used to pull the names from that list (males, or people in urban zip codes, or people who had spent more than $100 in the past year.)
The “creative” is the emailing itself. In the old days, that would include the envelope (windowed or solid, red or blue); in the email world it certainly includes the “subject line” of the email. Of course, it also includes the wording of the email and its presentation: headlines, graphics, etc. It includes the “from line”: does the email come from a person or the company? (Neal says it’s generally more effective if the email is signed by a person at the bottom, but that’s not necessarily what you want to show to get the email opened.) You can test a short email against a long one. And you can test the impact of HTML email against plain text.
The “offer” is what you are putting forward to the customer for an action. Historically, the action being sought was a purchase, and the offer might include a sweetener — a discount, for example. In an online context, it might be something else, like suggesting that the recipient come back with a request for a free chapter, or that the recipient offer their testimonial for the author or the book, or that the recipient register at a web site to stay in touch with the author or the publisher or the subject.
Neal argued strenuously that there should be a testing component to every mailing. He gave us a history lesson that dramatically underscored the advantages of the new email world. When direct response mailings were packages in envelopes, there was, of course, a significant cost for each piece dropped. Mailings would cost often 35 to 75 cents each — or, in marketing parlance, $350 to $750 per thousand pieces mailed — including the unavoidable cost of postage. For reasons that direct marketers could only hypothesize about but were borne out by experience, mailings tended to be most effective in January and July, with less activity occurring in between. And because the cost of mailing was so high, lists were rigorously “tested” before being “rolled out.”
But the process of mailing, getting response, and analyzing response was quite time-consuming. It took a minimum of weeks; sometimes it took months. Incorporating the results of the testing — to change the list, creative, or offer — would then occur six or 12 months later.
Emailings deliver results within days, if not hours. Reflecting what is learned from a mailing in a new mailing takes days, not weeks or months. That’s why Neal believes that the testing regime should be continuous.
Neal also taught us an old direct mail principle: TIOLI. That’s “take it or leave it” and it refers to the belief that there should always be a simple offer with one choice: usually buy or not. Now, this violated one of Shatzkin’s First Laws of sales, which is “always give the customer a choice of answers, one of which is not ‘no'”. “Red or green?” “One dozen or two dozen?” Of course, this law was formulated for a face-to-face selling situation (the kind I’m more familiar with). Neal saw the logic of that and agreed that direct response in the email world might enable a back-and-forth that would have been impossible in the snail mail world. His response was predictable.
“It’s testable,” he said.
Neal also wanted to make the point that just because emailing doesn’t cost any cash, that doesn’t mean that mailings are “free.” They’re not, because every emailing will result in some number of your names “unsubscribing”, or revoking the permission they had granted for you to mail to them. (We see this at The Shatzkin Files. Most of our posts result in some “unsubscribes” from our mailing list, although there are always more new subscribers than unsubscribes.) As Neal pointed out, all your names have a “lifetime value”: the average number of dollars you’d expect to get out of them in the fullness of time. If you mail to 100,000 names and 2% unsubscribe, you’ve lost their lifetime value. If they each had a lifetime value of $25, you’re losing 2% of that, or 50 cents per person on your list. That’s a “cost” of $500 per 1000 mailed!
In other words, don’t casually spam your lists! And rigorously measure the cost of your unsubscribes.
These two posts are really surface-scratchers, but they spell out the basics every publisher should take on board today. If a publisher isn’t already pulling together the names it has into a marketing database, creating a way to capture responses and sales information and tie them to the names, building their list of names creatively and aggressively every way they can, and beginning a program of rigorously-monitored communication with those names, they’re behind where they should be. They need a dedicated and focused effort monitored by top management.
These posts have not touched on many complex subjects publishers will face once they start this effort. As one of the comments on the first post has already said, Amazon has more consumer names, and more information about the books they want, than anybody. All of the other online booksellers will have theirs too. Authors are aware that they should own names and many big authors already have a lot of them. There are complex negotiations in everybody’s future about how those names can be used and shared to everybody’s advantage. A publisher will only be a factor in that conversation if it has its own well-structured and robust marketing database.
Bookstore shelfspace will continue to reduce. Publishers can’t stop that. Amazon will continue to dominate the business of selling books online to consumers and publishers can’t stop that. But publishers can stop the sheer waste of customer contact and outreach, which includes placing into many hands printed books that could carry URLs and mailing list solicitations and offers on them. It is past time to start building equity with what has always been thrown away.