Previous  |  Next


Jun 26

Peter Brantley

Peter Brantley

Where lies the Print on Demand

When On Demand Books, a Print on Demand (POD) device manufacturer, released news that the New York Public Library had installed an Espresso machine, a debate on POD business models broke out on a small mailing list that I run. With the permission of the participants, the debate is reproduced below. I think it well reflects many of the key issues of distributing POD as a business model.

** Mike Shatzkin starts:

The question is not: "to POD, or not to POD". The question will be "centralized POD, or distributed POD." I think distributed POD has its place, but I doubt very much that it's in the bookstores and libraries residing within fast shipping range of a strong centralized POD operation, for quite some time.

I don't know the relative cost of an Espresso-generated book versus a Lightning one. But let's remember that any bookstore or library can have any POD book in a day or two now. Once we get past the "wow" factor of having the book made on premises, the related questions of price and quality become determinants.

And before we get too carried away about what bookstores won't need to carry or pre-manufacture (either way creating risk), recall that it takes Espresso (at least) five minutes to make a book. Not too many bookstores can survive selling 12 books an hour, so this supplements, not substitutes for, inventory. (I'm sure its creators well understand that.) And customers may find the "printed here and now" feature a little less alluring if there are four people in front of them.

But if this DOES work more efficiently than I'm imagining, then why stop at bookstores and libraries? Wouldn't we expect to see one of these in every Kinko's and Staple's as well? If that were the case, great for the publishers and authors, but what would it do to the bookstores and libraries?

One case history worth recalling: a very large book retailer gave up its own POD operation several years ago and turned the activity over to Lightning. The idea was, "we'll take it back in a few years." A few years later, Lightning had added another trim size to its line, plus hard-backing and color capabilities. So the retailer, now having sufficient volume to support the prior capabilities, doesn't have the volume to support the new ones. So they aren't taking it back now, as they once thought they might, and Lightning keeps building new capabilities with added volume.

What will the Espresso-owning store think when they find they are ordering half their POD books from elsewhere because they can't handle the size or other spec requirements AND paying more for the ones they make themselves than they'd pay to have it delivered?

** Jim Lichtenberg says:

Excellent Points. Three thoughts:

First, I don't think this system was created for bestsellers or even front list books, but rather for slower moving titles from the backlist that probably wouldn't find shelf space in most retail contexts. More like: " in 5 minutes," vs., "delivered within 24 hours." (Anyway, Mike, only New Yorkers are so freaked out about standing in line).

Second: PoD is part of the life-cycle management of a title, which would be transparent to the customer.

Third: PoD is a distribution mechanism, and as such, Espresso probably has an expanding role to play as future miniaturization makes it possible to print a book with a book-sized machine ... well, maybe not that small. Just as we see certain books sold in non-bookstore venues, as time goes by, we will probably find on-demand book-printing systems in equally interesting places.

A report that I prepared last year on POD (available from Vista International, now Publishing Technology) includes a case study of Cambridge University Press' PoD/Short Run printing program, which makes the following point:

In summary, within Cambridge digital printing is not considered a different business, (although revenues are identifiable.) Rather than being seen as just a "printing" program, it has come to be viewed as a different variant of the supply chain. It is not really about printing, but rather about delivering books in a profitable way, and making product available to marketplace via a mechanism that allows the Press to achieve financially acceptable results.

** Mike Shatzkin says:

These points are accurate, but don't address the central- vs. distributed-POD paradigm that concerns me.

To narrow the question from his first point, ALL the Espresso machine does is reduce the delivery time from 24 hours to 5 minutes (if there is no line) - not from days or weeks or months. That's an advantage, but it is a LIMITED advantage. Within the same numbered point, Jim reminds us - accurately - that this capability is really for deep backlist, not for fast-moving new titles. So we are gaining this defined (and limited) time advantage only on a list of titles that is also defined, and though not so limited, generally constitutes books that can generally be waited for (if the wait is 24 hours.) Remember that PDFs delivered to the consumer's own computer can also address the rare cases of great urgency at much less expense (though admittedly with much less of a "wow" factor!)

Distributed POD will have to deliver incremental margin compared to centralized POD to gain widespread acceptance. When you factor in the cost of the machine, the space to place it, the training to use it, the waste from screw-ups, and the management of components (paper plus), then we've given the distributed-POD model some significant hurdles to climb to match just ordering from Ingram (Lightning) for profitability.

** Jason Epstein says:

The Espresso machine does much more than reduce the delivery time from days to minutes. It compresses the existing supply chain to the time it takes to transmit a digital file to the point of delivery, thus reducing inventory and delivery costs to consumers and publishers, and radically decentralizing the world-wide marketplace for books so that readers without access to traditional sources can have the same access to deep backlist as clients of the New York Public Library, and greater access in the digital future than readers anywhere enjoy today.

Beta versions of the Espresso have been operating cheaply and efficiently for several months in The Alexandrina Library in Egypt and the World Bank Infoshop in Washington DC with only the minor glitches typical of prototype machines. A later commercial version of the machine now in development will print, bind and trim library quality paperback books quietly for a penny a page and can produce a 300 page book in about five minutes with minimal human intervention in infinite trim sizes between 4.5x4.5 and 8.5x11.

Cost of inventory, storage, delivery returns, etc. will be eliminated so retail prices can be reduced, and returns to publishers and authors can be increased. Unlike existing POD technology, the Espresso will not require factory placement or skilled operators, and will cost about as much as an office copier and occupy as little space. Readers will be able to order titles from their home computers or mobile phones and pick their books up from nearby locations at their convenience, or have them delivered that day by ground transport without waiting in line.

The Espresso is designed to operate 24/7. The machines will be leased, not sold, to booksellers and libraries. They might also be placed on cruise ships, at hotels, coffee shops, supermarkets and so on, markets which conventional POD cannot serve. Broader reach, faster service, and lower costs and prices combine to produce a compelling business case.

The Espresso is not intended for current hardcover best sellers, but for decentralized delivery of deep paperback backlist, the sine qua non of civilization that is imperfectly served by today's technology.

** Joe Esposito concludes:

I don't really have a point of view on this topic, but I want to note that the options are greater than the central vs. distributed POD paradigm; they include "pure" ebooks and "extreme" distribution to the edge of the network, meaning an end-user's personal workspace. Hewlett Packard and Xerox (not to mention Kodak, with its huge retail presence) will certainly be watching as costs come down for POD; at some point POD will move to the desktop, just as color laser printing has.

Of course, if Espresso can get 5,000-10,000 machines out there, the number of corporate buyers, beginning with Amazon, is large. The POD winner will be the ace marketer. This is no longer a technical game.

tags:   | comments: 10   | Sphere It


0 TrackBacks

TrackBack URL for this entry:


But in going POD in bookstores and/or libraries you seem to have missed another major offshoot:

Books not on shelves. You'll have to be hunting a book to find it. It won't be on the shelf. As POD rises, what will happen? Will we get more and more best sellers ONLY on shelves and everything else, a lesser risk, delegated to the ease of POD? Will people find them then? How? What does that to the backlist except expand and marginalize it even more. If a book is backlist but people get an interest and ask for it a few times, more will be ordered, it will get shelf space and more people will see it. With POD that simply /will not happen/.

So how does that play into things? Should it?

This is without going into publisher issues with POD and contracts and writer's issues. Just one slice of the pie not addressed here.

bowerbird said:

kinko's already has machines that can make books,
-- or could be easily adapted so that they could --
especially if we aren't hung up on glossy covers.
and since they have franchises in so many places,
it'd be _really_ stupid not to use their ubiquity...
(and if they don't get smarter about it really soon,
some other copying franchise will eat their lunch.)

and why all of the focus on major publishers?

haven't y'all gotten the memo? we no longer care
about the publishing houses. we do peer-to-peer.
bookstores just introduce an artificial "scarcity"...

so most of the time we'll just use _e-books_...

when we _do_ want hardcopy -- and we will --
the reader will paypal the profit to the author,
who will transmit the file to a kinko's near to
the reader, and the reader will pick it up there
and pay kinko's for the printing and binding...

but yes, yes, yes, these machines will also be in
libraries. and yes, yes, libraries _will_ still exist,
because there are lots of reasons why they must,
most notably because they're information centers,
and information is _increasingly_important_ to us.

and the reason these machines will be in libraries
is because it's cheaper to print a book on demand
(and give the publisher their cut) than to shelve it.
and the cost curves will diverge more in the future.
(most libraries are _already_ busting at the seams.)

besides, it just doesn't make one whiff of sense to
print a book in one place and then pay the costs of
_transporting_ it, when we could transport the bits
for _free_ and print the p-book at its destination...


monopole said:

As a bibliophile and not a bookseller, my take on the time/distribution would be an overnight/two days lag. More specifically, if an online bookseller offered a reasonably priced guaranteed service which would print, bind and ship from a nearby regional distribution hub via priority mail with a tracking # I'd be overjoyed. In the same sense if I could make a click and mortar purchase involving an online request coupled with easy pickup at a local store I'd be even more overjoyed.

The critical issue isn't delivery in minutes, but predictable guaranteed delivery in a day or two, with a sliding cost level. When I'd like to have a book, such as one that I'd read for pleasure, I'm willing to wait a while for it, as long as I know I will get it, and receiving it is convenient. On the other hand, when I need a book for work or for a special occasion, I usually need it as soon as possible, preferably that night or at worst in two days, and I need a guarantee that I will get it.

Amazon fails horrifically in both respects. I've personally run into horror stories where a supposedly "in stock" and very expensive technical book will be delayed for months and then shipped overnight at horrific expense. In the same manner, "in stock" books will be delayed by months and then announced to be unobtainable. Pre-ordered titles will be shipped weeks after the release date. As a result, Amazon has become a source of absolute last resort for me. Given this corporate culture, POD is DOA at Amazon.

On the other hand, a smart and agile competitor could really wipe up the floor with Amazon by taking a page from early fedex tactics. The capacity to provide guaranteed service would be a massive plus. For click and mortar operations this could involve a kiosk that would allow for automatic pickup of pre-ordered books, allowing one to skip the inevitable lines and cluelessness of the clerks at large bookstores. Finally, just imagine that instead of a blank stare and "get it at amazon" a bookstore clerk answered "come back in an hour and we'll have it".

In addition, POD has the advantage of flexibility in terms of printing. The capacity to incorporate a bookplate with the owner's name in the pages as well as the cover would be very desirable. In the same manner a special dedication within a gift book would be equally valuable. Flexible printing options such as enlarged fonts and the like would be a major selling point as well. Don lancaster has written extensively on value added content at

Thomas Lord said:

Bowerbird and Monopole, commenting above, have excellent ideas.

One key thing to emphasize, to make the discussion better rounded, is that in between distributed and centralized production are a lot of intermediates -- "regional" production for some definition of "regional". Bowerbird shows an example with the concept of printing from Kinko's though, surely that is not the only viable form of regional production.

Regarding brick-and-morter bookstores: it seems to me that their main (and not insignificant) value-add is, of course, editorial selection in collusion with their customers. That is: a brick store is a good place to go to find what it was predictable you'd want as well as to find what good sense suggests you might want. Perhaps regionalized production can give that value-add a boost by absolutely smashing away a lot of transaction costs that currently limit what is on the shelf in a given store.

Tim and other O'Reilly folks have kindly explained, in previous posts, the trickiness of their having to meta-manage the on-shelf inventories of retailers. O'Reilly might figure out, or might even scout out, that a given store isn't moving a certain title because, through an inventory management glitch (e.g, theft) it has disappeared from the retail shelves. But it's silly expensive to notice that happening and it's stupid expensive to fix it. If regional production of books, including a very efficient "recycling" program were to spread, it might be possible to fine tune the shelf inventory of retailers in a much finer grained way. In other words, it isn't just "print on demand" but also "print on speculative demand".

Speaking of inventory management: a *huge* potential advantage of distributed production of books is the opportunity to standardize book formats (from the perspective of one retailer) in ways that centralized production won't do in a gazillion years. For example, a retailer might discover huge inventory management efficiencies if they are allowed to make the novel assumption that most titles on the shelves are in a format that includes a bar-code on the *spine* of the book so that the inventory of an entire shelf might be taken, quickly, with a simple laser scan. It would take eons for publishers to agree to such a marking of books yet a regional printer would have no problem creating such standards, even for works from different publishing houses.

In any event: decentralized (whether regional or fully distributed) production will (must) become the norm, simply for ecological reasons. Long haul shipping is, of necessity, becoming outrageously expensive for finished goods. It's a lot cheaper to maximize efficiencies of scale by long-hauling raw materials and long-hauling bits, postponing final processing and assembly as far as possible.


Jim Rait said:

If we skip over to the customer/consumer experience and plot the jouney of the person through the library/bookshop/coffeeshop then we may get a different perspective on the "user experience" If I remember my wife's visits to the NYP Library then she had to wait for books to be pulled up from the basement for her to browse; having determined which ones were worthy of a read with a close scrute Marilyn asks for them to be kept for the next day... soon we return to the UK so the book link breaks... imagine if one day while Marilyn takes an espresso break the book gets printed off for retention? Or round the corner at Borders, near Scriveners? They only have one of each book on the shelf and while we drink coffee our choices are run off? Tom Peters said many years ago that we should imagine the day when Borders announces it is stopping stocking books as the coffeeshop business is so profitable? Is this the day coming closer?
And then imagine O'Reilly's coffee house where we go and browse... like chapter 1 of this and chapter 3 of that...etc order an espresso with this chap1 and that chap3 etc. Drink and collect the hybrid book and pay.... Makes you think?

Some thoughts engendered by the responses so far.

1. As for what's on shelves and what's POD in a library, let's remember that once the library has a printed copy -- whether ordered from a publisher or created POD -- it will end up on the shelf. So only the FIRST copy of an obscure book would have to be "found" without a view of it on the shelf.

2. An enormous tool for inventory management in both libraries and bookstores is about to make its presence felt in our supply chain: RFID. RFID eliminates such things as requiring the bar code on the spine for taking stock or knowing whether a book ordered and shelved but not sold has disappeared through theft. Granted, full implementation of RFID is certainly years away, but it is also almost certainly coming. My colleague on this post, Jim Lichtenberg, is the industry's expert on this subject.

3. Borders could cut back on books in favor of more coffee. Does Starbucks put in one of Jason's Expresso machines now? (I think not, because five minutes for a product would be an eternity at Starbucks, but there's a nice symmetry to the speculation...)

4. Standardizing book formats is a constant struggle within many publishing houses. And once a publisher has decided on a trim size, it isn't always that easy, for design reasons, for a POD operation to change it by more than fractions of inches.

5. I was really struck by Monopole's post suggesting that Amazon routinely misses "promise dates" for book delivery. Between ebook reading and my very local B&N, I don't buy many books from them but this report is contrary to my understanding. Is it really true that Amazon gives bad service insofar as meeting expectations? I'd love to hear others on that...

Tho said:

2. An enormous tool for inventory management in both libraries and bookstores is about to make its presence felt in our supply chain: RFID. RFID eliminates such things as requiring the bar code on the spine for taking stock or knowing whether a book ordered and shelved but not sold has disappeared through theft. Granted, full implementation of RFID is certainly years away, but it is also almost certainly coming. My colleague on this post, Jim Lichtenberg, is the industry's expert on this subject.

That's trippy. Thanks for the trade news and, yeah, I should have known that. So, we're going to wind up in a situation where the fixtures in a retail store give the operator a continuous, real-time inventory -- very exciting.

There's a danger in this trend. From JIT to RFID, we're optimizing the hell out of the efficiency of retail without keeping up in production. So hyper-efficient retail keeps demanding cheaper and cheaper production (e.g., Walmart) which puts regional retail out of business and winds up in globalism for finished goods (with all of its exploitative consequences). We have to push hard on making manufacturing cheaper through technology and new markets in raw inputs, because as long as don't, manufacturing will just keep getting cheaper by lowering labor costs. There's only so much than can be done at the monetary policy level by allowing currencies to float, an so forth -- we also need the technical innovations for more regionalized production.

~ender said:

From the book-owner perspective, I *CANNOT* wait until we get real POD. However, it must be coupled with my ownership of the bits, in a non-proprietary format. Btw: leasing, renting, fee-paying are not the same as ownership.

When those two conditions are met, I will liquidate most of my (ridiculously huge) *physical* book collection. I like reading physical books, however they're a pain to store, organize, clean, etc.

If I can send a file to a POD center, that's not too far away, and go and pick up a book - then I'll be *set*. Anything I want to read in 15-30m? And I can just street-corner donate/recycle it when I'm done? Awesome. I'm probably a market for POD-in-the-home as well, depending on the price-point.

But costs also have to commesurate. I'm paying on average .75 a paperback, and between $1-$4 per hardback (non-technical) - by managing the used bookstores and other low-cost venues. $3 minimium for production costs for a paperback (tack on at least a $1 for profit for everybody) isn't very attractive. However, for the ability of getting some books (rare, hard to find in the secondary market), I'd be willing to pay that cost - my watch list is huge right now.

With a real file format, the limitations of size that Mike Shatzkin alludes to in #4 will not be an issue. If you've properly set out the text; auto-formatting will handle any size font, any size margins, and spare space for beginnings and ending of chapters, frontspieces, etc. Now if all you've done is make a PDF, or an image of the text - yes you will be so limited in size constraints. But only morons/anal IP owners are discussing that type of POD content.

Adam P. Knave's worry about the lack of browsing in finding new titles isn't a problem. This same problem occured with the elimination of the card catalogue. And in fact, is more pronounced with the CC's demise, as you still bump into simliar titles on library shelves when attempting to locate your search-engine generated call number - but you won't always get the cross-references in an electronic search, nor can you ever be sure that your ES included the whole physical catalog. In any case, with the advent of large databases of user-preferences (and datamining), we now have tools which can churn up recommendations of books to read/browse. Yes, many of these sites have been shutdown, and many have been taken over by profit-oriented businesses, but within another 10 years we'll have gotten it together - kinda like bluetooth is almost up to a working standard.

A deep backlist still has me drooling, but if POD wants more than one-time sales, they're going to have to free up the IP laws.

If they don't free up the IP laws, then POD books will have to meet or beat the quality of pre-published (what is the correct term?) and supply-chain distributed books. If I'm going to buy and own my bits in physical format (because I can't own (not rent!) them in byte format), then I'm going to wait until I can get it in a physical format that will last at least 20 years, and preferably until my death/my children's death.

And yes, does suck.

Physical shipping costs *are* high, and one of the reasons I do not order books online. Insurance covers the cost of the item, and not the shipping costs. Typically shipping costs are 3x the cost of the books I'm interested in. I've only found one site which will aggregate several used-book requests and match them to sellers in order to start minimizing those costs. That site could use a lot of work. I'd be interested in posting a large list, and seeing which sellers have the most books of that list for the lowest price, in aggregate. But that's not the case yet...

Paul Knecht said:

Are Border Bookstores No longer having booksignings for Print On Demand Authors, for example, I`m with Tate Publishing, I`m an author, my book wasd released March of 2007. I`ve been given a song and dance by a Mgr of a bookstore. My book does not qualify, since it is a print on demand. However the book is stocked Internationally. What vcan you tell me,who ever I`m writing to? Thankyou

Sam said:

I've been envisioning this in a science fiction kind of way for the last year or two, so it was great to read about the ATM POD model reaching a level of possibility.

Two comments: What quality is the paper/binding? Is it archive quality, or meant to be read once and discarded? Given time, cheaper versions of these machines will eventually make it onto the market, printing books with the cheapest materials available.

And, what happens to browsability? Knave commented on that right off. My thought: By that time, we'll all be at the library in Second Life (or something like it) anyway. You put on the helmet, wander around the virtual library, put books in you 'cart,' take off your helmet, and take your basket of real books home, freshly printed.

Leave a comment

Type the characters you see in the picture above.

Subscribe to this Site

Radar RSS feed