vendor-managed inventory

The future of books in stores


The future for books in retail stores is not unified; it’s dispersed. To the extent that there continue to be bookstores (and although shelf space in them will continue to decline inexorably, they’ll also be around for years to come), the bookstores will increasingly be more about books for reading and less about books for using. Much of the slack can be picked up by merchants of other things, but there are challenges.

The one piece of good news from Barnes & Noble’s most recent reporting was that their stores are still throwing off cash. We don’t know how much of the margin they’re reporting comes from books as opposed to NOOKs or toys and games. But it is definitely good news that the stores, which publishers still depend heavily on for sales as well as “discovery” are apparently still healthy.

Unfortunately, it would be a surprise if things stayed that way for very long. The share of book sales that are migrating to the Internet keeps growing. Amazon’s print book sales keep going up (more slowly, of course, but everybody else’s are going down) and sales of ebooks keep rising as more and more people get the digital habit. Amazon gets 60% or more of those sales through the Kindle platform.

The term “showrooming” is becoming familiar to people in the book business to describe the retailer’s role in the consumer’s new tendency to use stores to shop but the Internet to buy. Part of what drives this effect is Amazon’s well-earned reputation for heavy discounting — it seems just about every book at Amazon sells for less than the publisher’s suggested retail and most books in stores sell at the price the publisher printed on the book. (It was actually instructive in a recent NY Times piece decrying the reduction of discounting at Amazon to see that even academic books with very limited audiences were being sold at some discount from the publishers’ suggested price.)

With handheld devices that can check Amazon (or any other online) prices now ubiquitous, capitalizing on showrooming isn’t surprising consumer behavior, but it keeps bookstore retailers from “capturing all the value they create” as their own revenue.

Ultimately, illustrated books and the publishers who create them will be the most affected by these changes. There are two important reasons for that. One is that “straight text”, narrative books that are read from beginning to end work just fine as ebooks. That means they’re already cheaper and it is easy for more and more consumers to purchase them this way. (And from the publisher’s perspective, their margin is — at least for the moment — fully replaced when a sale migrates from print to digital.) The other reason is that a novel doesn’t need to be seen or touched to be considered for purchase. Even with the capability to “look” or “search” inside the book, many illustrated book customers really want to examine the printed version to make a buying decision. As there are fewer stores carrying them, that gets harder and harder for the consumer to do.

One of the changes we’re living through is that content as a “pure play” is getting less and less viable at retail. For Amazon, “media” (i.e. content) is no more than 20% of their business. It’s what got them started and it is still very valuable because the content people search for and buy sometimes can provide important clues about what else you can sell them. At least some of Amazon’s success against online media competitors is due to the fact that their base is broader than media.

Selling media alone has become a dinosaur in brick-and-mortar. Stores selling music and renting video have all but disappeared. Retail shelf space for books isn’t ever precisely measured, but what’s available in book-centric stores must be less than half of what it was five years ago, when Borders was still in business and before cutbacks in shelf space that are visible in Barnes & Noble and others. One of the hopes for traditional publishers is that smaller independent stores will pick up some of the slack. But the kind of stores they’re envisioning would probably carry less in the way of illustrated books, particularly illustrated how-to books.

All of this should spell opportunity for other retailers, particularly those who are in “verticals” where there is a lot of publishing: gardening, home repair, and crafts, as examples. Just about every retailer could benefit from a customized selection of books that would both attract and excite their core audience, often stimulating them to buy the other things the store sells.

But doing that is hard because buying books is hard for all retailers to do but it is particularly challenging for non-book retailers. They get foiled by the unique characteristic of the book business that frustrates just about everybody coming into it from the outside: its sheer granularity. A store that wants to carry 100 or 500 SKUs on gardening, home repair, or crafts will most likely need books from several, perhaps dozens, of publishers to have the best selection. And they’d be selecting from 10 or 100 times as many titles as they want to carry. New titles will be issued every week. Each individual title might have a sales potential in any one store of $150 or $250 or $500 at retail, less than any other single item that store has ever carried or thought of carrying.

The biggest publishers of illustrated books in the categories that can benefit from non-book merchants are all quite aware of their importance to their future. If you talk to people at companies like Abrams, Chronicle, Quarto, and Workman — and I have — they will all tell you that “special sales”, the industry term for sales outside the bookstore trade, are critical to their future.

Of course, publishers have been doing special sales for many years, certainly including the five decades that I’ve been involved in the business. They have done it in ways that aren’t necessarily optimal. They’ve forced stores to “buy”: select the titles and quantities and place orders for each shipment they get. That’s an unacknowledged bottleneck. It has also engendered two sales policies which are counterproductive but well-established. Special sales accounts customarily buy from publishers at high discounts (lower costs) than bookstores but, unlike bookstores, don’t get the rights to “return” unsold stock. This has “taught” some publishers that returns aren’t “necessary”; retailers should just mark down what they can’t sell.

And it has taught the retailers to expect unrealistic margins. Of course, those margins are also largely unrealized, because they are buying stock without the right to return and end up marking down a too high (but unknown to the publisher) percentage of what they buy.

Of course, stores that don’t return any other merchandise don’t know anything’s missing in their terms. But, ultimately, it reduces those stores’ ability to experiment and it reduces the publishers’ ability to get stock in place on speculation. They can only sell “sure things”, and even those end up not being sure things.

But that’s not the biggest constraint. The challenges of mastering the mechanics of buying are. Non-book retailers simply don’t have the inventory management systems or the ordering practices that are necessary to manage books, where good practice might be to bring in one copy 20 times over a year to get 20 sales. Why? Because the book might only sell 1 or 2 or 5, and putting in 20 to sell 20 would result in overstocks most of the time.

There is a better way for distribution to work for non-book retailers, and that’s with vendor-managed inventory, relieving the retailer of the need to manage complexity challenges greater than they face in their core business for what amounts to a sideline. So far, we are only aware of one distributor — West Broadway Book Distribution — that offers that capability. (Full disclosure: West Broadway is our client, and we had a lot to do with creating their offer and their system over a decade ago.) WBBD gathers books from many publishers for their retailer clients. That’s also almost always necessary because very few publishers have enough titles in any category to stock a store adequately on their own.

The future of bookstores is challenged. The likelihood is that those that survive will be smaller (or, like today’s B&N, devoting some of their floor space to things other than books). The book-centric retailer will be increasingly inclined to stock “writerly” books rather than “practical” ones. That creates an enormous opportunity for non-book retailers to create a traffic magnet, incremental margin, and a stimulus for their customers to buy their principle lines of merchandise by creating book departments. More of them will, but the challenges of buying will continue to be a constraint in the market.

15 Comments »

Vendor-managed inventory: why it is more important than ever


The idea of vendor-managed inventory has never become particularly popular in the book business, despite a few experiments over the years where it was implemented with great success. (And despite the fact that I was pushing for it back in 1997 and 1998.) But as the book business overall declines, with the print book business leading the slide and that portion of the print book business which takes place in retail stores falling off at an alarming rate, it is time for the industry to think about it again.

In fact, VMI for the book business began with the ID wholesalers and mass-market paperbacks right after World War II. The IDs — the initials stood for “independent distributors” — managed the distribution of magazines and newspapers at newsstands and other accounts within their geographical territory. The retailers had no interest in deciding how many copies of LIFE they got in relation to Ladies Home Journal; the ID made that determination. And since only the torn off covers were necessary for confirmation of a “return”, the “bulk” cost of distribution was in putting the copies in, not taking back the overage. And because newspapers and magazines had a disciplined frequency, it was obvious that you had to clear out yesterday’s, or last week’s, or last month’s to make room for the next issue.

When the first mass-market paperback publishers started their activity right after World War II, providing books for, among others, returning servicemen who had access to special servicemen’s editions of paperbacks (in a program created by the polymath Philip Van Doren Stern, a Civil War historian and friend of my father’s) they helped the jobbers along by having monthly lists. They also were comfortable with a book only having a one-month shelf life and having the stripped covers serve as evidence the book hadn’t been sold.

For quite some time, the initial allocations to the ID wholesalers (the local rack jobbers were called “Independent Distributors”) were really determined by the paperback publishers. Eventually, that freedom to put books into distribution choked the system, but there were a lot of other causes of the bloat. By the 1960s, many bookstores were carrying paperbacks and many other big outlets were served “direct” by the publishers, leaving the IDs with the least productive accounts. But VMI, even without any system and very little in the way of restraints on the publishers, was responsible for the explosive growth of mass-market paperbacks in the two decades following World War II.

In the late 1950s, Leonard Shatzkin, my father, introduced The Doubleday Merchandising Plan, which was VMI for bookstores on Doubleday books. For stores that agreed to the plan, reps reported the store’s inventory back to headquarters of Doubleday books rather than sending an order. Then a team posted the inventories, calculated the sales, and followed rules to generate an order of books to the store. Sales mushroomed, particularly of the backlist, and returns and cost of sales plummeted. Doubleday was launched into the top tier of publishing companies.

In a much more modest way, a distributor that my father owned called Two Continents introduced a VMI plan in the 1970s. Even with a very thin list and no cachet, we (I was the Marketing Director) were able to get 500 stores on the Plan in a year. We achieved similarly dramatic results, but from a much more modest base.

Two Continents was undone by the loss of some distribution clients. The Doubleday plan was undermined by reps who convinced headquarters years after my father left that their stores would be more comfortable if they wrote the Plan orders rather than letting them be calculated at headquarters. And the rise of computerized record-keeping systems for inventory and national wholesalers who could replenish stock quickly improved inventory performance, and store profitability, without VMI. Although our client West Broadway Book Distribution has successfully operated VMI in specialty retail for more than a decade, and Random House has worked some version of VMI at Barnes & Noble for the past several years, the technique has hardly been considered by the book trade for a long time.

It is time for that to change. What can foster the change is a recognition about VMI that is readily apparent in West Broadway’s implementations in non-bookstores, but would not have been so obvious to the bookstores using Doubleday’s or Two Continents’ services.

From the publisher’s perspective, the requirement that there be a title-by-title, book-by-book buying function in the store in order for the store to stock books purely and simply reduces the number of stores that can stock books. The removal of that barrier was the key achievement of the ID wholesalers racking paperbacks after World War II. Suddenly there were thousands of points of sale that didn’t require a buyer.

From the store’s perspective, buying — and managing the supply chain to support the buying decisions — is expensive. VERY expensive. Books are hard to buy. New ones are coming all the time; the number of publishers from which they come (and who are the primary sources of information about the books, even if you could “source” them from wholesalers at a slight margin sacrifice for operational simplicity) is huge; the shelf life of any particular title is undeterminable; and the sales in any one outlet are very hard to read.

Consider this data provided by a friend who owns a pretty substantial bookstore.

Looking at the store’s records for a month, 65% of the units sold were singles: one copy of a title. Only 35% were of books that sold 2 or more. (I didn’t ask the question, but that would suggest that 80-90 percent of the titles that sold any copies sold only one.)

Then, the following month, once again 65% of the units sold were singles. But only 20-30 percent of them were the same books as had sold as singles the prior month. Upwards of 70% of them were different titles. And upwards of 70% of the ones that sold one the prior month didn’t sell at all.

To further underscore how slowly book inventory moves, another report they do shows that more than 80% of the titles in the store do not sell a single copy in any particular month. So it is no surprise that an analysis of books from a major publisher that promotes heavily showed that more than half the new titles they receive from that publisher don’t sell a single copy within a month of their arrival in the store, which would include the promotion around publication date!

These data points demonstrate another compelling reason for VMI. When a store sells none of 80% of its titles in a month, and of the ones they do sell 80% of those sell one unit, they clearly need information about what is going on in other stores to know which ones to keep or reorder and which ones to return. Above the Treeline is an inventory service which provides its stores with broader sales data to address that issue, but the information is not as granular or as susceptible to analysis as what a publisher or aggregator could do with VMI.

Partly because of the high cost of buying and a supporting supply chain that a book outlet requires, publishers will see shelf space for books drop faster than retail demand. (The closure of Borders, which wiped out a big portion of the shelf space, is part of what is behind the recent good sales reports from many independents.) At the same time, retailers of all things will be under increased pressure to find more sales as the Internet — often, but not always, Amazon — keeps eating into their market.

This all adds up to VMI to me. We’ll see over the next couple of years whether industry players come to the same conclusion.

9 Comments »

Some ideas for publishers that will help bookstores; other suggestions that make us skeptical


This is the fourth of a series of posts on bookstores and their future. The previous posts have covered the challenges of buying (proposing VMI as a possible solution), explored what we should expect for the future of Barnes & Noble, and envisioned what the world of brick-and-mortar book retail might look like in the years to come. I promised previously to review the list of suggestions for publishers to help bookstores recently rounded up by Bookseller editor Philip Jones. He’s written more about this since, but the “original” list from Philip included:

1. Publishers offering books to retailers on consignment. That means the store pays when the book sells rather than on a date based on when it was shipped to them.

2. Publishers offering books to retailers with higher discounts. That means giving stores more margin between the price they pay and the price the publisher “suggests” as the retail price.

3. Bookstores taking advantage of Amazon’s “weaknesses” as an online bookseller. That would apparently be about localized curation as opposed to algorithmically-based suggestions.

4. Bookstores becoming something more (or less, but different) than bookstores. This suggestion may be inspired by B&N’s claim that they are creating new “prototype” stores.

5. Publishers creating special print editions for stores. This was done in Canada by the device of Random House creating Indigo-specific editions for Canada’s biggest bookstore chain.

And since then, in a radio interview, Harper UK MD Victoria Barnsley added a sixth suggestion:

6. That bookshops should charge “admission” to allow browsing (and perhaps credit the admission charge against a book purchase.)

We’re going to dismiss suggestions 3 through 6 pretty quickly. They either don’t scale or don’t help.

The notion that indie stores can beat Amazon at online selling is nothing short of preposterous. What indie stores can do, and should do, is offer an online sales capability to allow the customers they have who want to express their loyalty to do their online shopping with them. And they should do that in the simplest and easiest way possible. To the extent that the store has done curation work (store bestseller lists, recommendations from staff or customers), those should certainly be reflected online. But the notion that a single player can beat an online behemoth at the behemoth’s own game is a delusion and no great effort should be wasted on it.

The idea that bookstores become something other than bookstores, which is how I’d interpret suggestion number 4, is also not really much help. If not a “bookstore”, what, exactly? And if you can’t tell somebody “what, exactly”, then how is this advice anything more than a suggestion to keep throwing stuff at a wall until something sticks? That’s a strategy? Bookstores have already and always been “community gathering centers”. Playing up that piece of it is never a bad idea, but it hardly seems like an original one.

Similarly, the idea that publishers can save stores by offering them “unique” product is not really a solution at all. Yes, Random House (the biggest trade publisher) can do it for Indigo (a dominant retailer that owns the Canadian market). Even if it is adding value for Indigo, and we really don’t know if it is, there are precious few situations in the world where it could be applied.

And the suggestion that stores can save themselves by charging admission is one that can be very rapidly be disproven by any store that cares to try. It actually strikes me as a very good Candid Camera sequence. Put a toll booth at the front door of a retail store (any retail store, but a bookstore will do) and record the reaction of customers when they encounter something that makes absolutely no sense to them. I suspect wild enthusiasm for the idea will be rare.

However, the first two suggestions — to provide the stores inventory with more time to pay (consignment or extended payment terms) or more margin to work with — are worthy of more analysis and thought.

For publishers to consider easing the financial burden for bookstores based on their importance as a marketing component of the supply chain is a reasonable idea. But neither expanding retail discounts nor applying consignment is without complications.

Expanding margin needs to be done carefully, so that the margin expansion accomplishes the purpose that publishers seek: to increase the display of books in retail stores. Simply increasing discount is a difficult way to do that. What needs to be applied is an expansion of an existing principle.

In the book business, “coop” is the heading under which publishers purchase display for their books in prime locations. Coop was originally used for publishers to purchase space for their titles within a local bookstore’s newspaper ads. (Sometimes that “local bookstore” was a branch or group of branches of a chain.) But recently it has been applied to getting prime display locations, often near the cash register, as part of a promtion. The convention is for the payment for the space to be calculated as a percentage of a “supporting order”. This process imitates what happens in other classes of trade and is referred to outside the book business as RDA (retail display allowance) or MDF (marketing development funds). Another application of the same idea is for publishers to pay for “pockets” (sometimes called “slotting fees”). Under an arrangement like that a non-book retailer (like Michael’s, the craft store chain) can get an additional subsidy over and above what the discount schedule calls for on every title they carry.

But what we may be learning is that every book in a bookstore, or perhaps any retail location, has its discovery enhanced, not just the ones on promotional tables. So perhaps a publisher (followed by others, in time) might consider extending the idea to pay a “shelving fee” for every book in a “qualifying” bookstore. Consider a little math.

Let’s imagine a store that does $2 million in annual sales. If their average discount is 40% (which is a reasonable number; discount schedules would say it is higher than that, but it is reduced by freight costs, including for returns), the value of the inventory to make those sales is $1.2 million at cost. If they turn their stock three times a year, the average cost value of the inventory in the store is one-third of the total, or $400,000.

The two million in annual sales means shifting about 133,000 books (if the average retail price of the books is $15), and the average inventory is about 45,000 books. If publishers paid ten cents per book per month to be shelved, that would deliver an additional $4500 a month — $54,000 a year — to the store. If publishers paid 25 cents per book per month to be shelved, the store would get an additonal $135,000. Since a bookstore would be doing quite well to earn 10% on its sales, our notional $2 million store would be happy to earn $200,000 in profits now so, in either case, the “shelving fee” would be adding a meaningful increment. Certainly, for some stores it could make the difference between staying open or closing down. For others, it would encourage a bigger book inventory. In either case, that’s what publishers want to accomplish.

Publishers could, if they chose, make the “shelving fee” applicable whether the store bought the book directly from them or from a wholesaler. It actually makes it less tricky to apply if the wholesaler-supplied books are included. Invoicing now is done when publishers ship books, not when they arrive at the store so the time lag in between works in the publishers’ favor. For a “shelving fee”, publishers wouldn’t want to pay for time the book is not on the shelf: while it is in transit, or in a box waiting to be unpacked, or in a stockroom unavailable to a browsing customer.

In order to collect “shelving fees”, a store would have to deliver much more robust data than they now have to publishers about stocking and selling. But modern technology can make doing that not terribly difficult (systems don’t routinely do it now, but they surely could) and, in fact, stores should want to know about the efficiency of their shelving practices for their own reasons. And doing things this way would put publishers and stores on the same side around returns, because both would have good reason to get books that can’t sell off the shelves (and replace them with ones that have better odds).

Increasing the margin as a reward for a brick-and-mortar store being open and stocking books is doable and it is doable without cutting the wholesalers out of the picture. Consignment is definitely more complicated. And perhaps less helpful.

Sometimes the sale-and-return convention that has prevailed for nearly a century in the US book business is thought of as equivalent to consignment, but it isn’t. Although bookstores sometimes use returns as a tool to diminish the payments they have to make to publishers, they also “own” (and, in many cases, have paid for) a lot of books on their shelves at any particular time. And a non-trivial side effect of sale-and-return is that “shrinkage”, books that don’t sell but for whatever other reason may disappear from a store, are very much the store’s problem, not the publisher’s.

Under consignment, the payment from stores to publishers would be based on what passed through the cash register, not what was shipped from the publisher’s (or wholesaler’s) warehouse. “Shrinkage” would only be detected if a publisher called for a return of a book it had previously shipped and the store was unable to send it. Since even with the best of intentions, a store wouldn’t necessarily know a book was missing and certainly couldn’t pull a missing book for a return, the payments for those books would, at the very least, have to wait until some inventory check or returns protocol was invoked and discovered it.

The big question in consignment is when and how often a store pays. I recall having a discussion about consignment with a very large book retailer ten years ago. The top person there was thinking in terms of paying publishers every six months or so. It is safe to assume that no publisher would be excited about offering consignment on that basis. Allowing a store to “pay on sale” is one thing; allowing them to pay six months after sale is much more costly to the publisher.

For consignment to be workable, payments would have to be no less frequent than monthly, and would have to cover sales pretty much up to the moment of payment. What might make sense, for example, would be payments on the 5th of the month for sales made through the end of the preceeding month. That would be 35 days after sale for some books, 5 days after sale for others, and an average of about 15-20 days after sale. It wouldn’t be unreasonable for a publisher offering consignment to want payment more often than that, perhaps even as often as weekly.

The challenges of turning consignment into a workable commercial practice in our business include establishing a payment timing that makes sense and some method to catch shrinkage.

But the next problem is that the process of ordering would probably have to change. It is sometimes said that stores are now too easily tempted to over-order because, after all, they can return whatever they don’t sell. Imagine how much less restraint there would be on over-ordering if the store could hold books cost-free for as long as it took for them to sell! (There could still be the cost of freight in and out to discourage over-ordering, but that exists now.) Unlike the “shelving fee” concept, consignment puts the publisher and store in conflict around slow-moving inventory.

Let’s also take note of the fact that consignment is not all about paying later; sometimes consignment would require paying earlier. Bookstores get a boost when a bestseller comes in and flies off the shelves for the first week or two it is out. The revenue on those books is kept by the stores for 45 or 60 or 75 or 90 days (depending on how publishers enforce their collections) before they have to pay the publisher. Under a consignment arrangement, they’d have to turn over the publishers’ share much faster. (Of course, at the same time, they wouldn’t have to pay for some slower-moving books that might have come in the same shipment but hadn’t sold yet.)

There are other complications to consignment. The way things work now, publishers carry books in their warehouse on their balance sheet at inventory “cost” (something like manufacturing cost). When they sell them, they book the amount they sell to the store for, and keep some “reserve” for potential returns. On the store’s balance sheet, the books sit at the price the store paid, or will pay, the publisher for them.

But if the books are shipped to the stores on consignment, there has been no sale. So the publisher would have to continue to carry those consigned books on their balance sheet at the manufacturing cost and not credit themselves with the sale until the store reported it and paid them. What this would do to public reporting and bank covenants is a company-by-company proposition, and perhaps a knotty problem in some cases.

And sometimes there are state or local taxes based on “inventory”. IANAL (“I am not a lawyer”) but the taxing authority probably expects payment from the entity that ownsthe inventory. Under sale-and-returns, stores “own” it (whether or not they’ve paid the bill). Under consignment, the publisher certainly owns it. That would create complications, at the very least. Complications could also arise over insurance. (If a store had a flood or fire, would consigned inventory be covered by a store’s insurance?)

The bottom line is that publishers can help stores most by helping them carry their inventory less expensively and there are a great variety of ways to do that. The simplest way of all, of course, is just to extend the payment time from the current (as it often enforced) 60 days to something more. Thirty-five years ago, my father had me administer a program called “credit for overstock” where we gave stores 180-days extended billing for books left unsold after Christmas if they’d delay returning them. (Simple to do: issue a credit for what’s there dated today and an invoice for the same stock dated six months from now).

We’ve heard through the grapevine that at least one of the Big Six is experimenting with 180-day terms and that another might be a fast follower. That strategy is apparently offering competitive advantage (stores stock more of that publisher’s books, so they sell more of them too). That’s a way for a publisher to give benefits that are “like consignment” without the complications. From my perspective, it’s a shotgun, not a rifle, because it extends terms for everything equally. Credit-for-overstock targeted books that would very likely have been returned. The old “dated billing” plans targeted particular titles at particular times of year. Consignment requires that books that sell fast be paid for fast. A big across-the-board increase in time to pay is a far less targeted tool, but it still constitutes a big step in the right direction.

That’s because books on bookstore shelves are more valuable to publishers than books in their warehouse. Increasing recognition of that fact is occurring; more actions will certainly follow.

Worth mentioning — and inadvertently neglected by me in the VMI post — is that VMI does not need to be, and should not be, at odds with bookseller-management of curation. A publisher can certainly manage lists of titles that are designated “do not stock” or “always have on hand” that are designated by the store. The point to VMI is not to take tastemaking power away from the store. Of course a store should be able to exclude books they find offensive or that they think their customers will find offensive. And their decisions about categories or authors to stock out of proportion to how well they sell — higher or lower — can also be accommodated. And so can their inputs about local promotions that a publishers’ central office would have no way to know about. VMI offers two enormous benefits in any case. One is that the publisher knows things about individual book promotions, and recent performance, that might not be factored into each store’s calculations. And the other is that most stocking decisions are routine and  best made — particularly in the age when we’re discovering Big Data — by a system massaging the maximum amount of information.

25 Comments »