Vendor-managed inventory (VMI) is a supply paradigm for retailers by which the distributor makes the individual stocking decisions rather than having them determined by “orders” from an account. The most significant application of it for books was in the mass-market paperback business in its early days, when most of the books went through the magazine wholesalers to newsstands, drug stores, and other merchants that sold magazines. The way it worked, originally, was that mass-market publishers “allocated” copies to each of several hundred “independent distributors” (also known as I.D. wholesalers), who in turn allocated them to the accounts.
Nobody thought of this as “vendor-managed inventory”. It was actually described as “forced distribution”. And since there was no ongoing restocking component built into the thinking, that was the right way to frame it.
The net result was that copies of a title could appear in tens of thousands of individual locations without a publisher needing to show up at, or even ship to, each and every one.
To make this system functional at the beginning, the books, like magazines, had a predictable monthly cycle through the system. The copies that didn’t sell in their allotted time were destroyed, with covers returned to the publisher for credit.
Over time, the system became inefficient (the details of which are a story for another day, but the long story short is that publishers couldn’t resist the temptation to overload the system with more titles and copies than it could handle) and mass-market publishing evolved into something quite different which today, aside from mostly sticking to standard rack-sized books, works nothing like it did at the beginning.
My father, Leonard Shatzkin, introduced a much more sophisticated version of VMI for bookstores at Doubleday in 1957 called the Doubleday Merchandising Plan. In the Doubleday version, reps left the store with a count of the books on-hand rather than a purchase order. The store had agreed in advance to let Doubleday use that inventory count to calculate sales and determine what should then be shipped in. In 18 months, there were 800 stores on the Plan, Doubleday’s backlist sales had quadrupled and the cost of sales had quartered. VMI was much more efficient and productive — for Doubleday and for the stores — than the “normal” way of stocking was. That “normal” way — the store issues orders and the publisher then ships them — was described as “distribution by negotiation” by my father in his seminal book, “In Cold Type”, and it is still the way most books find their way to most retail shelves.
After my Dad left Doubleday in 1960, successor sales executives — who, frankly, didn’t really understand the power and value of what Dad had left them — allowed the system to atrophy. This started in a time-honored way, with reps appealing that some stores in their territory would rather just write their own backlist orders. Management conferred undue cred on the rep who managed the account and allowed exceptions. The exceptions, over time, became more prevalent than the real VMI and within a decade or so the enormous advantage of having hundreds of stores so efficiently stocked with backlist was gone.
And so, for the most part, VMI was gone from the book business by the mid-1970s. And, since then, there have been substantial improvements in the supply chain. PCs in stores that can manage vast amounts of data; powerful service offerings from the wholesalers (primarily Ingram and Baker & Taylor, but others too); information through services like Above the Treeline; and consolidation of the trade business at both ends so that the lion’s share of a store’s supply comes from a handful of major publishers and distributors (compared to my Dad’s day) and lots of the books go to a relatively smaller number of accounts have all combined to make the challenge of efficient inventory management for books at retail at least appear not to need the advantages of VMI the way it did 60 years ago.
And since so many bookstores not only really like to make the book-by-book stocking decisions, or at least to control them through the systems they have invested in and applying the title-specific knowledge they work hard to develop, there has been little motivation for publishers or wholesalers to invest in developing the capability to execute VMI.
Until recently. Now two factors are changing that.
One is that non-bookstore distribution of books is growing. And non-bookstores don’t have the same investments in book-specific inventory management and knowledge, let alone the emotional investments that make them want to decide what the books are, that bookstores do. Sometimes they just simply can’t do it: they don’t have the bandwidth or expertise to buy books.
And the other is that both of the two largest book chains, Barnes & Noble and Books-a-Million, are seeing virtue in transferring some of the stocking decisions to suppliers. B&N, at least, has been actively encouraging publishers to think about VMI for several years. These discussions have reportedly revolved around a concept similar to one the late Borders chain was trying a decade or more ago, finding “category captains” that know a subject well enough to relieve the chain of the need for broad knowledge of all the books that fall under that rubric.
This is compelling. Finding that you are managing business that could be made more efficient with a system to help you while at the same time some of your biggest accounts are asking for services that could benefit from the same automation are far more persuasive goads to pursue an idea than the more abstract notion that you could create a beneficial paradigm shift.
As a result, many publishing sales departments today are beginning to grapple with defining VMI, thinking about how to apply it, and confronting the questions around how it affects staffing, sales call patterns, and commercial terms. This interest is likely to grow. A well-designed VMI system for books (and buying one off-the-shelf that was not specifically designed for books is not a viable solution) will have applications and create opportunities all over the world. Since delivering books globally is an increasingly prevalent framework for business thinking, the case to invest in this capability gets easier to make in many places with each passing day.
VMI is a big subject and there’s a lot to know and think through about it. I’ve had the unusual — probably unique — opportunity to contemplate it with all its nuances for 50 years, thanks to my Dad’s visionary insight into the topic and a father-son relationship that included a lot of shop talk from my very early years. So here’s my starter list of conceptual points that I hope would be helpful to any publisher or retailer thinking about an approach to VMI.
1. Efficient and commercially viable VMI requires managing with rules, not with cases. Some of the current candidates to develop a VMI system have been drawn into it servicing planograms or spinner racks in non-book retailers. These restocking challenges are simpler than stocking a store because the title range is usually predetermined and confined and the restocking quantity is usually just one-for-one replenishment. We have found that even in those simple cases, the temptation to make individual decisions — swapping out titles or increasing or decreasing quantities in certain stores based on rates of movement — is hard to resist and rapidly adds complications that can rapidly overwhelm manual efforts to manage it.
2. VMI is based on data-informed shipments and returns. It must include returns, markdowns, or disposals to clear inventory. Putting books in quickly and efficiently to replace sold books is, indeed, the crux of VMI. But that alone is “necessary but not sufficient”. Most titles do not sell a single copy in most stores to which they are introduced. (This fact will surprise many people, but it is mathematically unavoidable and confirmed through data I have gotten from friends with retail data to query.) And many books will sell for a while and then stop, leaving copies behind. Any inventory management depending on VMI still requires periodic purging of excess inventory. That is, the publisher or distributor determining replenishment must also, from time to time, identify and deal with excess stock.
3. VMI sensibly combines with consignment and vendor-paid freight. The convention that books are invoiced to the account when they are shipped and that the store pays the shipping cost of returns (and frequently on incoming shipments as well) makes sense when the store holds the order book and decides what titles and quantities are coming in. But if the store isn’t deciding the titles and quantities, it obviously shouldn’t be held accountable for freight costs on returns; that would be license for the publisher or distributor to take unwise risks. The same is really true for the carrying cost of the inventory between receipt and sale. If the store’s deciding, it isn’t crazy for that to be their lookout. But if the publisher or distributor is deciding, then the inventory risk should be transferred to them. The simplest way to do that is for the commercial arrangement to shift so that the publisher offers consignment and freight paid both ways. The store should pay promptly — probably weekly — when the books are sold. (Publishers: before you get antsy about what all this means to your margins, read the post to the end.)
Aside from being fairer, commercially more logical, and an attractive proposition that should entice the store rather than a risky one that will discourage participation, this arrangement sets up a much more sensible framework for other discussions that need to take place. With publisher prices marked on all the books, it makes it clear to the retailer that s/he has a clear margin on every sale for the store to capture (or to offer as discounts to customers). And because the publisher is clearly taking all the inventory risk, it also makes it clear that the account must take responsibility for inventory “shrink” (books that disappear from the shelves without going through the cash register.)
Obviously, shrink is entirely the retailer’s problem in a sale-and-return arrangement; whatever they can’t return they will have paid for. But it is also obvious that retailers in consignment arrangements try to elide that responsibility. Publishers can’t allow a situation where the retailer has no incentive to make sure every book leaving the store goes through the sales scan first.
4. Frequent replenishment is a critical component of successful VMI. No system can avoid the reality that predicting book sales on a per-title-per-outlet basis is impossible to do with a high degree of accuracy. The best antidote to this challenge is to ship frequently, which allows lower quantities without lost sales because new copies replace sold copies with little delay. The vendor-paid freight is a real restraint because freight costs go down as shipments rise, but it should be the only limitation on shipment frequency, assuming the sales information is reported electronically on a daily basis as it should be. The publisher or distributor should always be itching to ship as frequently as an order large enough to provide tolerable picking and freight costs can be assembled. The retailer needs to be encouraged, or helped, to enable restocking quickly and as frequently as cost-efficient shipments will allow.
5. If a store has no costs of inventory — either investment or freight — its only cost is the real estate the goods require. GMROII — gross margin return on inventory investment — is the best measurement of profitability for a retailer. With VMI, vendor-paid freight, and consignment, it is infinity. Therefore, profitable margins can be achieved with considerably less than the 40 to 50 percent discounts that have prevailed historically. How that will play out in negotiations is a case-by-case problem, but publishers should really understand GMROII and its implications for retail profitability so they fully comprehend what enormous financial advantages this new way of framing the commercial relationship give the retailer.
(The shift is not without its challenges for publishers to manage but what at first appears to be the biggest one — the delay in “recognizing” sales for the balance sheet — is actually much smaller than it might first appear. And that’s also a subject for another day.)
6. Actually, the store also saves the cost of buying, which is very expensive for books. The most important advantage VMI gives a publisher is removing the need for a buyer to get their books onto somebody’s shelves. The publisher with VMI overcomes what has been the insuperable barrier blocking them from many retail establishments: the store can’t bear the expense of the expertise and knowledge required to do the buying. It is harder to sell that advantage to existing book retailers who have invested in systems to enable buyers, even if some buyer time can be saved through the publisher’s or distributor’s efforts and expertise. But a non-book retailer looking for complementary merchandise that might also be a traffic builder will appreciate largely cost-free inventory that adds margin and will see profitability at margins considerably lower than the discounts publishers must provide today.
7. Within reasonable limits, the publisher or distributor should be happy to honor input from the retailer about books they want to carry. It is important to remember that most titles shipped to most stores don’t actually sell one single unit. Giving a store a title they’re requesting should have odds good enough to be worth the risk (although that will be proven true or not for each outlet by data over time). Taking the huge number of necessary decisions off a store’s hands is useful for everybody; it shouldn’t suggest their input is not relevant. Indeed, getting information from stores about price or topical promotions they are running, on books or other merchandise, and incorporating that into the rules around stocking books, will help any book supplier provide a better and more profitable service to its accounts. After all, having a store say “I’d like to sell this title for 20 percent off next week in a major promotion, would you mind sending me more copies?” opens up a conversation every publisher is happy to have.
Of course, in a variety of consulting assignments, we are working on this, including system design. It is staggering to contemplate how much more sophistication it is possible to build into the systems today than it was a decade-and-a-half ago when we last immersed ourselves in this. In the short run, a VMI-management system will provide a competitive edge, primarily because it will open up the opportunity to deliver to retail shelves that will simply not be accessible without it. That will lead to it becoming a requirement. As I’ve said here before, a prediction like that is not worth much without being attached to a time scale. I think we’ll see this cycle play out over the next ten years. That is: by 2025, just about all book distribution to retailers will be through a VMI system.