I contributed to a long-standing industry argument I usually try to avoid when I speculated that ebook growth could lead to a situation which threatened the returnability model for book inventory shipped to retailers and wholesalers. I should have been more emphatic that what I was actually suggesting was that the model of using speculatively-printed inventory to sell books was threatened, and that returnability, which is a subset of that model, would go along with it.
Coincidentally, Ken Auletta wrote a New Yorker piece at the same time in which he demonstrated that lots of smart people, he among them, don’t actually understand the economic impact of returns, let alone the promotional impact of the practice of using books as posters and display props, which is responsible for most of them.
Misunderstanding the economic implications of returns, failing to grasp how it is most useful to think about and analyze them, and various misinterpretations of them are very common up and down the ranks in publishing, in big companies and small.
I recall in the early 1980s I had a small publisher client that was distributed by a larger one. I used to go into the big publisher’s office every couple of weeks and leaf through the sales rep orders, trying to figure out which reps were best selling my client’s books effectively and which ones perhaps needed to get a refresher course on the sales handles.
The sales director in this shop, who ran sales departments for several large publishers in his career and who was both a nice guy and thought of as a “numbers man”, kept tabs on the returns percentage from the two national chains: at that time, Walden and Dalton. He calculated those percentages meticulously at the end of each month, based on that month’s shipments out and returns processed.
Well, you could count on the fact, every year, that returns percentages for both chains were astronomical in February, when few new big books shipped and every excess of optimism going into Christmas was punished. And you could also be sure that both chains had low returns in November and December, when retailers are much too busy building up stock for holiday sales to send anything back.
In other words, the calculation produced a result that was, literally, useless. It simply confirmed the obvious. I bring up this straw horse because it demonstrates an important fact that is just as true when analyzing returns for an account for any period of time, even a year.
The returns in any period of time are at least partly driven by purchases made in a previous period of time. So if sales in a prior period were high, the returns percentage in a subsequent period of lower sales will also be high, and that’s regardless of the appeal of the books being shipped in either period.
So as sales fluctuate, as they inevitably do, publishers will find that all weak sales years have apparently high returns and strong sales years have low returns. This isn’t cause and effect; it is more like a tautology: the inevitable consequence of the fact that returns are based on the sales made three months ago, six months ago, and sometimes a year ago, not on the sales being made right now.
One way a publisher might try to analyze their way around the timing problem is to look at returns by title which, if the calculation were done when an edition’s life was completed (perhaps on the hardcover after it has been remaindered deep into the paperback’s life), would seem to be a valid number to analyze on a stable base: the shipment of one particular book.
But even calculating things that way is not very useful as an analytical device. Returns come from the inventory in the pipeline when the book declines or dies. If the book has already sold for years in that edition, the base for the returns calculation (all books shipped) includes those from many printings long past. The copies in the pipeline at that point might only be 10% or 5% of the total the book has shipped in its lifetime, so the returns will, of course, come in under those percentages.
A publisher trying to manage inventory efficiency needs to be concerned with the books in her own warehouse and the books currently in the supply chain and subject to return. Those sold long ago are not part of that inventory. Taking 100,000 copies back on a book that sold a million or two million and calculating the returns percentage won’t produce any flashing caution lights, whereas taking back 25,000 of 35,000 pushed out on a new “make” book will produce a lot more internal scrutiny and hand-wringing in most houses. But a tighter focus on how to to manage the inventory actually in the supply chain in the face of declining sales would be much productive than an “analysis” that produces nothing but a historical observation. A publishing company can realistically make a lot more progress and save a lot more money figuring out how to reduce the 100,000 than the 25,000.
One factor that affects returns percentages and is not often considered by a publisher is the frequency with which an account orders. I’ve never had the opportunity to do the analysis, but I’d bet there is a very strong inverse relationship between the frequency with which an account orders a publisher’s books (whether direct or from wholesalers) and returns percentage. There are two interrelated reasons for this.
The obvious one is that the store that reorders is signaling that they’ve sold books, meaning that what is available to be returned as a total of what they’ve ordered is less, the effect we’ve noted above. But the other is that an account that orders less frequently is raising returns as a percentage in one of two other ways (depending on the book): they’re either over-ordering to prevent going out of stock or they’re failing to reorder when they are out and losing sales, which means that whatever is returned (of other titles, in this case) is calculated against a smaller sales base than should have been the case.
And all of this is very important to take on board right now because publishers are likely to be entering a sustained period of declining brick-and-mortar sales (which is the part of the inventory pipeline most likely to generate returns), caused in part by declining brick-and-mortar shelf space. That means that advance sales on new titles are going to be lower than before and the number of backlist titles being stocked is going to steadily decline at the same time. And that adds up to a higher returns percentage on a declining sales base.
It doesn’t take a rocket scientist to understand that if stores close or cut back their shelf space, they’ll be sending books back that will increase returns rates. But it will not be simple to separate out how the current strategies for inventory placement are working and to avoid having the “noise” of returns made because of contraction cloud those judgments.
The “good news” is that the old measurements (total returns in a year against total shipments out in a year) will deliver an exaggerated picture of how much current new title sales are declining (although they will be delivering a true picture of a very sad fiscal reality.) The bad news is that we’re going to start hearing about companies whose overall returns percentages have gone from the 20s to the 30s and then higher.
I wrote in the piece that I referred at the top that it looks like sales registered online — whether print or electronic — will be half the sales of new narrative text books published by 2012. When that number gets to 90%, there won’t be much of a premium on a publisher’s ability to understand, and thus to be able to manage, their returns with sensitivity and sound analysis. But between now we will be living through several years when it will be one of the critical skill sets for all publishers fighting to survive sea changes in the business.
For a few years earlier in the decade, we had a thriving little business called “Supply Chain Tracker” in which we took the sales and inventory data provided by major accounts and delivered publishers reports in Excel to help with analysis. We created what we believe were some critical metrics to watch: percentage of stock on hand sold, week by week in each account; percentage of each book’s stock in an account’s Distribution Center (if they had one); total inventory in the retail pipelines we could see and what percentage was selling through in a week (by title), and then the same for wholesale and distribution centers. Most of a publisher’s supply chain inventory is really visible today, if a publisher takes the time and care to look. Just glancing at each spreadsheet an account delivers for how things are going on top sellers was never sufficient; the price to be paid for such inattention in the future will only escalate.