The Shatzkin Files

On Amazon stores and publishers accepting standardization; two unrelated commentaries

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Share on RedditShare on StumbleUponEmail this to someone

When the “Amazon-opening-400-stores” rumor landed a week ago, many people were gobsmacked. It took me a minute to get past that, which also required getting past my firm conviction when they opened the Seattle store last year that it was an information-gathering exercise, not the opening move of a bigger retail play.

But, when you think it through, it not only doesn’t seem crazy that Amazon would open stores, it seems like an obviously compelling move.

Other retailers that started strictly online have opened retail locations, most notably the eyeglasses shop Warby Parker. (This New Yorker story mentions that. It also has an interesting disclaimer at the end because “Amazon Studios is producing a New Yorker series in partnership with Condé Nast Entertainment”. Wow.)

“Omni-channel”, which is really a new-fangled fancy term for selling both online and through a brick store, is the buzzword du jour of retailing. Actually, the online piece of that is the harder part and Amazon already had that licked.

Barnes & Noble “beat” Borders largely because they had a network of distribution centers that made stocking their retail locations extremely efficient. Amazon’s network of distribution centers is complicated because it isn’t just books, but they have many times the number of points of inventory storage as B&N. In fact, they have many times the number of storage points as B&N and Ingram and Baker & Taylor combined!

Amazon has tons of information that nobody else does that would inform their stocking decisions if they harnessed it. They know where searches are coming from for particular book titles or for generic needs, both geographically and psychographically. And they probably can detect early lifts for particular books faster than anybody else, simply because they have more data.

It is possible that if B&N and the indies had responded differently to Amazon Publishing, agreeing to stock the books rather than boycotting them, this could have played out differently. (No stronger argument could be made for the efficacy of that strategy than this post arguing that stores should stock Amazon titles to punish them because the returns would make them unprofitable! You can’t beat logic like that.) If the stores had stocked their titles, Amazon might have chosen to use their distribution center advantage to start wholesaling, rather than to support their own retail locations (as they appear to be doing).

But the determination of the brick retailers to boycott Amazon was spelled out loudly and clearly. So opening Amazon retail locations — as it increasingly appears they have every intention to do — has two strategic payoffs for them. One is that it gives them access to at least some brick-and-mortar retail locations for their publishing output, which otherwise they can only sell online. And the other is that it capitalizes on their distribution centers, delivering additional sales and margin for investments already made.

In a recent post, I suggested one specific way Amazon could get very disruptive if they had more than a handful of stores. There’s another. They are a tech company that likes to have computers make decisions that in other companies and in other times have been made by humans. I suspect they’ll figure out pretty fast that they will want to have some sort of vendor-managed inventory system to streamline and optimize the stocking decisions for what will almost certainly be a growing network of retail locations. (The part of a trade book person’s DNA that is most out-of-step with the digital age is that we like to make decisions case-by-case, rather than living with decisions made by rules we create. That’s the key to the second half of this post.)

Sophisticated but automated stocking and restocking decisions are not part of the toolkit at B&N or of any other retailer or wholesaler we know. Could that be the next battleground that Amazon retail stores create? That would certainly be disruptive, but at least in this corner of the world it would not be a surprise.


One mantra of the book publishing world is “every book is different”. We sometimes refer to that fact as reflecting the “granularity” of the book business compared to other kinds of consumer goods businesses or other media. Even if you think in terms of categories, there are just more of them in publishing than there are for other products or media.

Perhaps, then, it isn’t surprising that publishers are often inclined to encourage that uniqueness beyond where it is required. And, frankly, it is only required for editorial development and for targeting the marketing. The objective at every place in the value chain in between should be to standardize and, as much as possible, to treat many different books the same. That’s not a creative imperative; it is a commercial imperative.

My father first experienced the tension that this insight can create at Doubleday in the 1950s when he persuaded the company to standardize the trim sizes of their books for maximum printing efficiency. That didn’t require radical changes. It simply meant that books would be an eighth- or quarter-inch longer or shorter, wider or narrower. These were differences that were really not perceptible to most people, yet it was a real internal corporate battle to wrest control from designers who believed “every book is different” and that this mystery (or cookbook) had to be published as a 6 by 9 inch book while that one had to be 6-1/2 inches by 9.

In fact, the trivial differences in trim size were not important at all to the books’ chances of success. There were other decisions — the specific paper or type face among them — that also had no discernible commercial impact on each individual book but were, nonetheless, intentionally made book-by-book as though they did. In many houses, and (admittedly I’m saying this without any supporting data) probably more in smaller houses than larger ones, they still are. And that’s true even though whether the paper is 55 pound or 60 pound or the type face is Times Roman or Baskerville can’t be shown to have any impact at all on a book’s sales.

Now the University of North Carolina Press has been funded by the Mellon Foundation to put Dad’s theory to use in the university press and academic publishing world. They’ve created a service offering through their Longleaf distribution platform that takes the design, pre-press, production, and distribution burden off the hands of university press and academic publishers so they can focus on what makes them distinctive: the books they choose to publish and the skill with which they edit them.

This fits an industry reality I identified a couple of years ago that I called “unbundling”.

On one hand, UNC Press Director John Sherer reports real success, expecting to grow that part of their business by 50 percent in the coming year. But he also reports resistance by some presses who believe that making these design and production decisions adds so significantly to the “quality” of their output that they’re comfortable losing money doing it.

My own hunch is that many directors just don’t have the heart (or courage) to get rid of staff that, with all the best intentions and capabilities but without the advantages of technology and scale, provide them with no better than average quality at a much higher cost than they need to spend. This was a battle for Leonard Shatzkin when he fought it at Doubleday in the early 1950s and apparently it is still being fought hard six decades later.

Tweet about this on TwitterShare on LinkedInShare on FacebookShare on Google+Share on RedditShare on StumbleUponEmail this to someone

  Back to blog

  • InklingBooks

    My own hunch about Amazon’s stores has recently taken a darker turn. I used to live near Amazon’s first store and often shopped at University Village. The location has two traits that I suspect Amazon values highly.

    1. It’s a highly trafficked location in an affluent neighborhood. That means lots of buyers who might—but don’t necessarily have to—pay more for convenience over low price. Those are the book customers Amazon most covets but can’t necessarily win over by simply being the cheapest. Not used to waiting for anything, they’re the ones who still find traditional bookstores appealing.

    2. Major bookstore competitors are nearby. That’s a critical factor in Amazon’s scheme. In this case that’s the marvelous University Bookstore about a mile away. Keep in mind that Amazon’s purposes are NOT like those of Home Depot and Lowe’s, who often put their stores only a few blocks apart. Amazon doesn’t want to benefit from the drawing power of that other store, and it isn’t willing to live with that competition in the long-term.


    To explain, I’ll make up some numbers for that Amazon v. University Bookstore competition in Seattle. Assume that over a certain period, Amazon sells $100,000 worth of books, mostly popular books and bestsellers. Assume for the sake of argument that selling books in a store at the same price as their warehoused books means it costs Amazon $110,000 to sell that $100,000 worth of books. Does that mean that a store that’s losing Amazon money is a bad idea and will be abandoned? Not if Amazon is thinking what I think they’re thinking.

    Staff at University Bookstore have already gone on record warning that the Amazon store is skimming off the cream of their sales, meaning the ones that make fast-flowing and certain profits and enable the store to maintain its utterly marvelous stock of slower-moving, more serious titles. That $100,000 in sales has come from their most valuable source of income.

    Now suppose, just suppose, that in many locations around the country that Amazon store lands near a city’s largest full-service bookstore as as happening in Seattle. That means that the other store loses its principal source of income and is either driven out of business or forced to drastically reduce its size. That is what I believe Amazon is aiming at. That’s why it’s willing to build stores that lose money.

    Do you get what I’m saying? By losing $10,000 on that Seattle store, Amazon may be setting up a situation where University Bookstore either goes out of business or is forced to radically reduce its inventory. With that inventory reduction, people who would have bought from University Bookstore now find shopping online with Amazon more appealing. By losing $10,000 on its store, Amazon has now so weakened its competitors that it will gain more than enough online sales to recoup that loss and more.


    And, of course, in the long run having fewer other outlets for books also means Amazon can squeeze its suppliers—publishers and authors—still more tightly. Amazon already pays Kindle authors the worst royalties in the industry (35% at prices where Apple pays 70%). With even less competition, it will pay even less and make still more demands.

    I saw something quite similar happened when I was living in Seattle during the 1990s. A friend of mine was a purchasing agent for Boeing and she had come to hate her work. Why? Because her company was having her call up suppliers and say, in essence, “This is to notify you that while we have been paying you $100 for this part, we will now be paying $80.”

    How could Boeing managed to do that? Check your corporate history. Boeing had recent bought MacDonald-Douglas, the country’s only other builder of large airliners. Those aerospace companies were caught in a trap. They must sell to Boeing at the prices Boeing set or they didn’t sell at all.

    That’s the future that I suspect Amazon envisions for publishers and authors. It won’t be “My way or the highway” for them. It’s be “My way or bankruptcy.”

    Unfortunately, as I watch publishing flail about, I don’t see any effective, well-thought out resistance being mounted. At best, the major publishers seem to be engaging in holding operations, such as keeping the ebook prices of their bestsellers high, because Amazon dominates that market, so they sell more print copies, a market Amazon does not yet own.

    What they are doing a bit like what happened in Dunkirk during WWII and, as Churchill noted, “wars are not won by evacuations.” They’re won by well-thought-out, well-executed strategies.

    And no, I don’t think staff cuts count for much. Reducing the quality and diversity of books is precisely what Amazon wants. It wants to reduce books to a commodity, produced and sold at the lowest possible price. You’re not going to defeat Amazon by playing its game by its rules.

    • Emily

      This is fascinating insight, thank you.

    • You’re absolutely right that Amazon wins — and can put a store under — by taking away a relatively small percentage of highly profitable sales. That’s because they don’t depend on the store for their existence and their competitors do. They have now spanned the book value chain from author to reader in a way that nobody else has or can or will. Not every problem has an answer. Amazon will continue to get bigger and they will continue to do so at the expense of everybody else in the publishing value chain EXCEPT those who were shut out of it before. Those would be the indie authors who never got agents or deals. You call for “effective, well-thought out resistance”. Nobody could be against that. But unless somebody spells out what it is, nobody can execute on it either.

  • Steven Zacharius

    Two thoughts from me. Right now Amazon doesn’t publish print books unless it’s via POD I would presume. This opens up a new world to them, especially dealing with returns. More importantly though, if they open physical stores they now would have Nexus in those states and not only would it be a sales tax issue but the states would now demand a specified percentage of income tax from Amazon. I don’t know if Amazon is one of those corporations that has it’s tax base oversees, but having Nexus is a very big and costly deal. As I said somewhere else, we would be dancing for joy if someone else said they were opening 400 new bookstores. More outlets for books is a good thing for our business and for consumers as long as everyone is playing on a level playing field.

    • I’m not sure we can ever have a “level” playing field again, if we ever did. And with the number of DCs Amazon now has, I’m not sure how many states they can avoid “nexus” in. By the logic I described, it would only make sense for them to put stores very close to where they already have DCs anyway, so it might be a non-issue. Glad to have you back here commenting, Steve. Haven’t seen you for a while!

    • agclaymore

      Wouldn’t they stock the stores from their own DC’s, relying on their massive data mine to ensure they only pick hot titles for each store’s region? I’d be surprised if they simply started taking truckloads from publishers’ distributors.
      If they do choose based on their own sales data, they might not have so many returns to deal with.
      It’s going to be a foggy road ahead. If they carry only the faster selling titles from major publishers, I would imagine there might be a stiff fight for marketing dollars when new titles are being introduced.
      Many headaches ahead for publishers, but also some opportunities for the more nimble members of the breed.

      • Oh, of course they’d stock themselves. You’re on the right track with the stocking strategy, but they’ll figure out they can do things nobody else has, like systematic title rotation. Yes, of course they’ll sell prime real estate when they have enough prime real estate to support the negotiations. As for returns, they might be so minimal they’d handle it with markdowns. Depends on their merchandising strategies. Maximum sales without returns is lower than maximum sales with returns.

    • SpringfieldMH

      I don’t pretend to know how Amazon does or does not publish print books right now… but can’t image there’s anything to stop them from selectively employing more traditional higher volume printing arrangements when estimated demand/actual sales indicates that such makes economic sense.

      Amazon stores as “more outlets for books”… yes, provided they don’t opt to boycott select books or other publishers… same as some big chain and indie book stores are now apparently boycotting Amazon books…

      Interesting times.

      • Of course, Amazon has access to — and already employs for its own publishing program — all the same printers the publishers do.

        And if (maybe when) their store network gets to any consequential size, you can bet your bippy it will be used to favor some and disfavor others. That isn’t news. That’s retailer-supplier reality. There is always variation among the “relationships” leading to different treatment when the rubber meets the road.

      • Steven Zacharius

        Of course they could make printing arrangements to publish their books traditionally. But keep in mind that there’s an enormous difference between printing books where there there was a demand for the books previously was primarily based on very low eBook prices. And if they owned 400 stores you can just do a print run of 1000 copies and make money. A minimum mass market print run would be at least 5000 copies. But starting with returns is an entirely different business than they’re in now. They are used to 100% sales.

  • Peter Goodman

    I really would be curious to see whether typeface makes a difference. Sure, doing a book all in sans serif will impact readability, but that’s not entirely a design issue. Will consumers react all that differently to a book typeset in Garamond vs Minion or some other similar but newer and more expensive typeface?

    Book designers will say yes, and they will further say that while the impact of book design should be unobtrusive it is very real and operates at a subliminal level. Yet as this article suggests, is this anything more than an unprovable assertion by people with a vested interest in its truthiness?

    I think many of us in publishing are scared of the actual truth, that our designs provide fulfilling work but in many instances not a lot of value added. (Cover design is something else, for now.)

    We work a lot with Lightning Source on shortrun digital printing. The LSI size format matrix rewards us for working with only a dozen or so standard trim sizes, like 5.5 x 8.5, 6×9, and some international sizes. So I’ve asked production to design books that at some point in their life will be headed to LSI to choose only from the available sizes. In most cases, this results in a change of no more than half an inch to what the designer might have preferred anyway.

    There was predictable blowback, and even I felt bad about it, like I was selling out our artistic standards somehow. And I still feel that way a bit, although for the life of me I can’t come with a really good reason not to do it. Shouldn’t a good designer be defined as one capable of handling a standard format and making magic with it?

    • I don’t mean to say that typeface makes NO difference, and that san serif and serif faces can be used interchangeably. But designers don’t need to choose from an unlimited number of faces. Really.

      I remember when we needed to rush either Gary Hart’s or Frank Mankiewicz’s book through production in 1973. It was set by hot metal in a shop in Manhattan for speed and control. The shop had five cases for one font, three for the other two. So we chose the one where we could put five operators on simultaneously rather than one of the two where we could only have three. Practicality trumps artistic decisions, particularly for narrative books that are meant to be read.

  • agclaymore

    There’s a third strategic advantage. Their extensive data regarding what readers in specific markets want to read. Their brick and mortar stores would be carrying a far more targeted offering than your average book-store.
    More wheat and less chaff.

    • Absolutely right. In our digital marketing business now, we pick up signals like that from other web sources to try to help publishers geo-target advance title laydowns. But Amazon has the best data of that kind that there is. And it isn’t available to anybody else.

  • Nathan Maharaj

    Agree with all you’ve said about Amazon’s strategic advantage in opening physical stores.

    But I’m not seeing nearly enough discussion about the retail innovations they’ve packed into this move. For example, the single innovation of having customers scan products to get prices is huge: for the first time a retailer can measure conversion in a bricks & mortar store the same way they can measure it online. And moves like this show how willing Amazon is to try retail as no other retailer has done it — which is how they got to where they are online.

    Books are a beachhead. Again.

    Target and Wal Mart should be very concerned.

    • It’s the same “secret”. They don’t need to make the margins others do because they have multiple ways to win. And they also seem to recognize that they win by learning. It proved very hard for bookstores to compete against them. The same will be true for Target and Walmart. Neither of them has cloud computing capabilities to peddle.

    • Steven Zacharius

      I don’t see having to scan each book to get the price as an advantage. Too slow.

      • Nathan Maharaj

        But if you’re Amazon the consumer already believes you have the lowest price–and the product is in their hands. Price is not the point: conversion data is. Reveals not only price tolerance but also registers that this specific customer visited this store at this time and expressed interest in this thing. That’s invaluable data for a retailer.

      • Steven Zacharius

        On many items Amazon is not the lowest prices especially if it’s included in prime. We don’t know how much money Amazon makes in their book business. It’s obviously not their biggest piece of the pie or anywhere near the most profitable, if profitable at all. They lump it in with other media categories in their reporting. The consumer would have to want to go into the Amazon store which would probably have a far smaller selection than a typical chain store or indie store.

      • If Amazon builds 400 stores, or even 100, some consumers will go to them because they’re the closest choice, or the only practical choice. I have been told by Amazon executives that they have never “lost money” on their book sales overall. Of course, they could never have built the service capabilities they have with that margin strategy except that their book business is housed within a much larger play.

      • Steven Zacharius

        There’s a big difference between losing money overall and operating a bookstore as you know. A bookstore is going to have a limited number of books while online it’s the long tail scenario that makes them so successful. And when it comes to them saying they don’t lose money we have no idea what they’re including. Do they include just the distribution business, the self publishing and/or Amazon Publishing. Amazon is used to running a sell through percentage in the high 90% range from their warehouses. It will be totally different even using technology when you have to keep many multiple copies of a book on the shelf and can’t ship it from another warehouse when the consumer wants it right away.
        Like I said though, opening 100 or 400 stores is good for publishers and readers whether it comes from Amazon, B&N, Books a Million, or some other book outlet. This business has gone from having thousands of different accounts down ten major accounts and a handful of successful indie bookstores. It’s really a shame for everyone but especially local communities who use them as a meeting place and a source of tax revenue. So bring them on and let’s have more books available by more writers.

      • For what it is worth, I was given the very clear impression from executives high up in their book business that they made money reselling books and ebooks from publishers. I’d interpret that to mean that they took in more dollars in sales than they spent paying the publishers for the books. That info is two or three years old.

  • bbmm18

    I see Amazon ‘ s next moves as the next phase of what happened in the ’90s with indies as well as the continued ongoing change to superstore chains. Though the playing field is lamentably not level, I’m not sure this is not so different from other national chains dominating an industry (including books). As well as other historic changes to the marketplace throughout the ages and especially in the last 150 years.

    I believe they’ll go for the high end demographics, and possibly a few high profile stores that may be more accessible to underserved neighborhoods. Many indies also are in or draw on affluent consumers and or areas with higher ed institutions nearby.

    In my comment on an earlier post I mentioned vans and distribution centers. The area where I see the Amazon vans used to have both a Borders and B&N. Now it’s got the Apple Store, a luxury movie theatre, high end supermarket, Bloomingdale’s moved up the road to a larger location. When you drive through the area, the Range Rovers appear out of nowhere–in recent months joined by the Amazon vans. No coincidence. And it’s not just books. It is easier to order everything. Back in the day Sears Roebuck even delivered components of a two-story home. What goes around . . .

    To wrap up–I love well designed print products but for most, cost and utility can win out, which is ok. In my production days I recall having to come up with solutions, one resulting in a book where on reprint we scrapped the spot uv on the cover, nice cover, still very attractive–changed binding from concealed wire – o to perfect. You can’t always make that binding change but it worked in this case. Sprial could have been a mid-level compromise. Poorly designed indie ebooks, cover and interior require improvement or standardization. Barbara Miller

    • I think what is different about Amazon today versus the B&N and Borders build-outs in the 1990s is that Amazon doesn’t have to make the same ROI on each store to win. If they capture customers they didn’t have before and weaken competitors so that more business ends up going online, increase their leverage with suppliers, *and *give exposure to their own books which enables them to sign up more books, they do just fine even if any store, or even the total of all the stores, isn’t quite profitable. B&N and Borders had a lot of Wall Street money to build out their superstore networks, but they didn’t have all those advantages.

  • bbmm18

    They never had to make money on books and for the most part, have not.

    • I don’t think they “haven’t made money on books”. They’ve probably made some. A little. But not nearly enough to grow as they have, to build the tech that they have, to gain the market position that they have. They’ve changed the game and it really isn’t possible for most of the incumbents in the business to change it themselves to match them.

  • Kira Gale

    Is it a possibility that Amazon will place 3 or 4 POD printers in their stores and gave customers the choice of all their POD titles? The customers could browse on internet screens looking for specific titles. How long would the print and bind process take? What if the book could also contain a personal inscription page, from a choice of formats? In the meanwhile, the customer could be buying drinks and snacks from a store cafe while waiting.

    • Wouldn’t work very well. Those in-store machines aren’t small. And the in-store machines don’t have the format flexibility that the industrial strength machines have. And, in fact, even if publishers cooperated, there would be lots of books that wouldn’t work because they’re too big or too fat or require color or high-quality picture (halftone) printing.