There is a vision of online bookselling, which I share, which is that it will become increasingly atomized. Books (and, ultimately, other content too) will be merchandised in unique ways across countless web sites curating and presenting content choices for their own communities and audiences. One early prototype of how this might work is the Random House initiative powering “bookstores” for Politico and Publishers Lunch’s Bookateria.
This is not a new idea. I remember a meeting more than five years ago hosted by O’Reilly Media in New York City to plan the first Tools of Change conference at which Brian Murray of HarperCollins, not yet their CEO, talked about how a way should be found to merchandise books on current affairs topics around and adjacent to today’s news stories that were relevant. The Random House capability, among many other things it can do, readily enables just that.
This is not necessarily bad news for the biggest online retailers like Amazon, B&N, Apple, and Kobo. The Random House execution delivers “their” customers to one of the others to consummate the sale and they’re rewarded for having pushed the “discovery” by collecting referral fees from the etailer which processes the sale. (How the revenue is split between Random House and the web site providing the screen real estate is not known to me, and presumably only one of a number of moving parts in the negotiations between them.) Doing things this way allows both Random House and their clients to avoid the two biggest (and closely-related) headaches of online bookselling: managing DRM and customer service. In addition, the costs for what is called “card and cart” – handling credit cards and providing shopping cart technology — are also avoided by handing off the actual transactions.
Bookish, the new discovery engine and bookseller which was financed by three of the Big Six, also offers referrals in addition to their own fulfillment (which is provided by Baker & Taylor).
Peter Hildick-Smith of Codex, our go-to guy for understanding the concept of “discovery”, says that bookstores offer discovery combined with availability, a “twofer”. In effect, web sites offering ebooks (and possibly print too) alongside their information and conversation are doing the same thing.
In fact, the same approach makes sense in the brick-and-mortar world, but it is a lot harder to do.
Merchandising is the bottleneck for any retailer, online or in stores, trying to sell books. Which books do you offer? Which books do you feature? What do you discount? This is a challenge online, which is why Random House believes it can build a business helping web sites do it. But it is even more challenging in a physical environment, which requires actual printed books to be displayed, sometimes to be sold and sometimes to be returned.
But smaller and more targeted displays of print books in stores — whether a general selection or one targeted to store’s other customers — also make more sense than big book superstores in the digital era. Physical bookselling locations can offer consumers convenience and speed. If you’re shopping, you can see more titles faster than you can online and you can walk away with your purchase rather than waiting for delivery.
Publishers gain access to their audience through retailers. Non-book retailers, just like web sites, are specialized in some way and they both attract and serve customers if they offer appropriate books.
The challenge for non-book retailers who would like to carry books is stocking them. Almost no matter what a store sells, from clothes to hardware to specialty food, there would be a selection of books that would please their customers and perhaps increase their sales of core items. This is obvious in, say, a crafts store or hardware store where just about everything that’s sold is part of a project (selections of which and instructions for which are often found in books) and could require instruction about how to use it most effectively (also content well suited to books).
Picking the right books is hard work. If the retailer buys them from publishers (whose sales representatives would know their content and could actually guide one to the best title choices for one’s audience), it is a hopelessly fragmented challenge. In many areas, you might find 25 good books that could require you to buy from 10 or more different publishers. The publishers’ sales terms will be one problem (minimum order sizes) and the administrative costs would be far too big to justify considering the small sales the store would get from ancillary merchandise like this. Wholesalers have the books of many publishers, but their teams don’t have the kind of title-level knowledge the store needs to make the selections.
Meanwhile, bookstores labor under a similar constraint. We pointed out in our recent B&N analysis that the cost of their supply chain gets harder to bear as sales of books diminish. Independent bookstores have also always been constrained by the cost of buying, although they don’t really see it that way because it is part of the landscape.
The core point is this: the responsibility for getting the right books onto retail shelves is one that has always belonged to the retailer. That reality encouraged, even required, large book retailing operations: big independent stores and large chains could amortize that cost across far more sales than a small bookstore or a little book department in another retailer.
There is one established way to reduce those costs: vendor-managed inventory. With VMI, the cost of negotiation — of conversation between a “buyer” and a “sales rep” — plummets. In addition, it is actually easier to stock the right books at the right time. A key component of making better decisions is making more decisions that cover shorter prediction times. Ordering more frequently makes it much easier to avoid over-ordering as a protection against going out of stock. That increases stock turn (the key to bookstore profitability) and reduces the need for returns (leaving more margin for both the retailer and the publisher).
As I’ve written previously, a long-standing client of mine called West Broadway Book Distribution has been operating a VMI system in a small number of non-book retailers for a decade. They have a system which interprets the sales reporting and makes restocking decisions based on them automatically. They also have a system to test new titles in a sample of a chain’s outlets to decide whether or not to roll them out. Their automation has enabled them to manage a lot of granularity — thousands of potential titles in more than a thousand stores with the books coming from more than a hundred publishers — profitably and with workable margins for both the retailers and the book-providing publishers.
West Broadway started because its owner had a few books of their own that they wanted to sell to a couple of “women’s hobby” accounts where they already had relationships. We encouraged them to be more ambitious and they were willing to try. So they aggregated the books from many of their competitors, larger and smaller, to add to their own and invested in the VMI system (which they might not have needed to make sales of their own books alone).
That’s a path we should expect to see other specialty publishers taking in the future. Subject-specific knowledge is helpful in doing that (although it can be done successfully without it).
Stocking a general interest store with VMI is much more complicated and will take more time to evolve. But bookstores can take steps in the right direction by consolidating their buying to a smaller number of suppliers and pushing all their really small vendor ordering to a wholesaler (or two) to gain efficiencies from managing fewer vendors.
Remember that one of the keys to efficient stocking is frequent ordering. Bookstores mostly understand that and order from wholesalers every day. But they probably also order directly from dozens of publishers. They do that to gain a little bit of additional margin and, perhaps, to reward the sales rep that calls on them to present the list.
I’m going to say flatly that the margin differential is almost certainly not worth pursuing for what it costs in stock turn (capital tied up) and risk (returns because people buy more copies when they’re tempted by the higher margin order). My father made that clear in numerous examples in his monograph, The Mathematics of Bookselling.
The rep reward is a little more complicated but most publishers these days figure out how to pay their reps for sales that go through the wholesalers.
Any store routinely dealing directly with more than 20 publishers and distributors will almost certainly improve their financial performance by cutting that back and consolidating. They might lose a little margin; they might miss a couple of smaller-potential titles (but not big ones), but their lives will be simplified and that will save a lot of money.
And with daily ordering from wholesalers, which just about all stores do, it becomes unnecessary to carry more than a copy or two of most books, except for the purpose of display prominence.
Once a bookstore has taken those steps, it is in a position to start demanding some VMI help, even if just from the sales reps. This was an idea that was pioneered in around 1980-81 by an indie in Shaker Heights, OH, called Under Cover Books in a project on which I consulted.
We were too far ahead of our time (the computers were too klunky), but the idea was that we gave the reps reports of how their titles were performing: on-hand, shipments in, and sales. Then they had an inventory ceiling stipulated and were free to order more books, of their choosing, up to the inventory ceiling. We then calculated the inventory’s performance (beyond the scope of this piece to get into that particular detail, but essentially combined the impacts of discount and turn) and raised the inventory level for the most profitable publishers and reduced it for the less profitable.
What defeated us was the complexity of administration. Part of that was because there were so many more smaller publishers then. Part of it was that the only way to communicate the inventory data was by shipping spreadsheets by snail mail (slow and not cheap).
This would be infinitely easier to do today, and the ease would be multiplied if you were only trying to do it with a handful of big suppliers.
I am only aware of one publisher today that has worked corporately on a VMI system for books, and that’s Random House. I believe they initially developed the capability and implemented it for chains: first for Barnes & Noble and more recently for Books-a-Million. But they also seem already to be prepared to offer the service to independents. Since, when the Penguin merger is complete in a few months, stores will be able to get damn near half the most commercial books from Penguin Random House, having “just” them operating VMI would constitute a sharp reduction of the store’s operational demands.
Whether or not this is what they’re thinking at the moment, the new Penguin Random House is bound to find it sensible to employ its VMI capabilities in self-defense to open retail print book outlets in places that are bereft of bookstores in the years to come. Those outlets will have space for shelves, customers and cash registers, but no ability to discern what books they ought to stock or what the timing should be of ordering. They’ll be sought out as necessary because bookstores, which are carrying the requirement of making these stocking decisions, will have increasingly become uneconomic (and therefore defunct).
This vision of the future is of books being sold mostly in stores that aren’t bookstores, enabled by VMI systems that largely don’t exist yet. It would be even better if the VMI vision took hold in time to save some of the bookstores that exist today to survive to that future time when the demands on them to manage inventory will have been ameliorated by necessity.
In my last post, I cited a bunch of suggestions pulled together by Philip Jones for how publishers could help bookstores survive and promised to review them. This post was intended to get to that, but I couldn’t get there within a reasonable number of words. Next time.