Archives for category: Book manufacturing

Another force pushing us towards a single world market in the book business may be identified in the internationalization of book manufacturing and distribution. Now of course publishers have been ordering printing from overseas for years, especially since World War II. High quality color? Go to Italy. Price problems? Or lots of hand work — add-ons? Hong Kong here we come. When I was first in New York, working for Cambridge University Press, it was quite obvious that the tidal flow of manufacturing was governed by £/$ exchange rate fluctuations: now it would be cheaper in the USA, then it would switch back to favor Britain. Work would follow: publishers are always looking for the cheapest way to make their books. The development of digital printing, primarily of print on demand manufacturing, accelerated these options. One of the incidental implications of print on demand is that you can minimize shipping costs by printing a book as close to the ultimate customer as possible. This cost saving may not look like much when you think of one book, but over a year it mounts up to a significant sum.

Local territorial markets be damned! Ingram‘s Vice President of Content Acquisition, Kelly Gallagher, tells us “Sometimes as much as 30 or 40 percent of a publisher’s Ingram wholesale can be to non-US addresses. That’s also very glass half-empty, half-full, because it’s a pressure point for the local retailers and distributors in those countries who used to have a corner on getting books into their market.” From a Publishing Perspectives interview.

Amazon lies at the heart of all this. Amazon’s distribution system is amazingly slick. They fulfill an order in the most efficient (cost and timing) method possible. There’s a cascade of options, which involves answering a series of questions including:

  • Are there several books in this order?
  • If so, where is each of them best sourced —
  • from our inventory,
  • which warehouse is that inventory in,
  • is it here or overseas,
  • or should it come from Ingram’s inventory,
  • from the publisher’s inventory,
  • from any other wholesaler’s inventory,
  • by using POD,
  • and if POD then POD at our own facilities
  • or at one of Lightning Source (part of Ingram)?

All all these steps the algorithm focusses on optimal proximity to the customer and speed of delivery. Amazon is heavily dependent on (or makes extensive use of) Ingram’s services. In many ways Ingram’s stock can be regarded as an extension of Amazon’s stock, as can all the print-on-demand files that Lightning Source maintains. The cascade, regardless of where the book ends up being shipped from, will always result in the books arriving on your doorstep in a grinning Amazon carton. Your order may be split into two or more separate packages — the cascade will have determined which is optimal.

Now this system is beyond impressive: Amazon can get most books to most people overnight. I sympathize with the old guard bookstores, but is it really right to go after a company for having developed a system that’s so efficient that almost everyone wants to use it? Doesn’t monopoly require a certain amount of unfair advantage: using your market power to destroy the opposition. Amazon is merely guilty of designing such an efficient system that nobody else can match it. Surely that shouldn’t be a crime, should it?

The classic would be McCain side sewing. This became the standard way of binding school books. An essential part of the educational process seems to be constant research by young pupils into new ways to destroy a book. A McCain-sewn book was virtually indestructible — without resorting to tearing the pages out, an approach which the ethics of this international research project seem to have ruled out.

The book block, obviously in need of jogging.

A McCain machine would start with a book block and drill holes vertically all the way through it near the spine folds and parallel to the spine edge. It would then stitch the whole thing together all the way down the spine edge, making it about as strong as it’s possible for a binding to be. Sure, you could get a knife and cut the threads, but anything approaching normal schoolroom wear and tear would fall to destroy the book.

For a thin pamphlet you could just run the thing through a Singer sewing machine, leaving a line of stitching parallel to the spine. Such side stitching might also be done using wire staples: see Binding styles 4.

This announcement from Printing Impressions caught my eye. Lay-flat has long been a bit of a bugaboo in the book world. Readers are alleged to want it, though most of us have actually failed to hear the voices and wonder how much extra people might be willing to pay to get it. You can see how something like a lab manual might need to lie open without the pages flipping back in the middle of your experiment so you have to dive for it and spill your beaker of acid all over the book and yourself. We used to deal with this demand by spiral binding or plastic comb binding the things.

Ota-Bind provided a methodology for lay-flat binding. Peleman Industries now introduces a machine, the V Twister Lay-Flat Paper Converter, which by bending a double fold back and forth allegedly breaks down the bonds between fibers in the paper at the fold allowing it to relax in the open position. Sounds good — the proof of course will only be known after someone serves up the pudding. It is an extra step, and unless you can integrate it into your binding line, would interrupt the binding process.

If you don’t see a video here, please click on the title of this post in order to view it in your browser.


A year ago, when coronavirus was just something we’d heard about happening in Wuhan, we were already wondering about capacity issues in book manufacturing. The business has been undergoing radical changes for decades, and the problems were coming to a head with plant closures and paper shortages.

Then on top of this came coronavirus lockdowns, and for a minute it looked like the whole shebang might fall apart: if there were to be no books to manufacture, then there’d be no manufacturers of books left. But not so fast: although publishers initially delayed a lot of books, after a short time we found out that we were actually able to keep on selling books despite all the problems we were facing. There were a few minutes there where States scratched their heads as to whether printing books was in fact “essential” or not, but that went away as we slid by on the basis that printing many things was obviously essential and it was deemed a good idea to allow all printers to keep on going in to work.

The book manufacturing industry could be said to be “suffering” from three interrelated long-term shifts —

  • a sharp reduction in print runs. Over the last five or ten years publishers have finally figured out (and print technology has enabled them) that printing fewer than a life-time’s supply of a book might not be altogether crazy. Demand planning and supply chain management are no longer just topics in books we publish. We are now reading those books.
  • a reduction in paper making capacity for book papers. You can probably make more money off making a less demanding grade of paper than book paper. Investments in paper making are immense, and tend to fall into cycles which lag the business cycle by a year or two. There are many fewer book paper manufacturers nowadays than when I started out in the business.
  • a technological changeover to digital from offset lithography. Just as lithography represents a productivity boost as against letterpress, so digital takes productivity further. But if you have a plant full of offset presses, you do face a difficult set of investment choices. Any technological change will take years to work its way through the industry as people continue to make money using fully-amortized “old” equipment.

I suspect there’s a pretty direct connection between these trends and the problems which have assailed LSC and Quad Graphics. In crude terms you could argue that such mega-companies were predicated on a regular supply of large print orders from big publishers. Obviously books like the Obama volumes still attract long print runs, but more and more books are being printed in shorter runs with more frequent reprints. Add to that the ease with which a publisher can now change printers: with a digital file as the start point for a printing, the up-front work on any printing is much simplified, and can in principle be done by any company. So if printer A who did the first printing can’t give you a good schedule, then off you go printer B. In the olden days, moving flats was a nightmare. Plants set up for the old dispensation of fewer and bigger printings are challenged.

One big change which coronavirus has brought about is a difference in the way a book is scaled out to the trade. Used to be you’d want to have stacks of books in every bookshop in the world on the day when you finally declared the book published. Doesn’t happen any more (at least not to the same degree) — if there aren’t crowds wandering into stores what’s the point of having piles of books there for them? Online ordering falls into a different pattern — perhaps not one we can define too precisely just yet. This must have a knock-on effect on the book manufacturers as demand for the product is extended over a longer timeframe. Further evidence of this change in sales patterns is provided by the frequent expressions of surprise at the greater and greater importance of back list sales.

Now there are those who believe that 2020-1 will represent the death knell of the print book. Such Jeremiahs will always leap to their favored conclusion. However I think what we have seen over the last year is just how  far people are willing to go to get a physical volume despite all the barriers social distancing impose. I hope I’m not guilty of leaping to my favored conclusion when I suggest that this scarcely seems evidence of people finally concluding that ebooks are the only way in which books should be supplied.

Printing Impressions has an article with the thoughts of four book manufacturers.

In 2015 I held forth on this topic.

Trinity College Library has a copy of Lexicon Alchemiae (1612) from Sir Isaac Newton’s collection. Examination of the book revealed some pretty extensive insect damage.

Their blog has a description of the steps taken in making this book look almost new: well, look more like an undamaged old book. Painstakingly cut-out pieces of paper were inserted into the holes left by the beetles. It’s interesting to note that restoration doesn’t go as far as inking in the missing type eaten by the bugs. Please check out the post: it’s a fascinating story.

Books aren’t the only Newtoniana they’e got at Trinity. They have an apple tree too: not the very one, but a graft from the actual Flower of Kent apple tree Newton would have sat under when gravity hit him in the head. Atlas Obscura has a brief piece about it.

The unit cost, the cost of manufacturing a single copy of a book, consists of two elements: the fixed costs and the variable, running costs.

Fixed costs

The fixed costs are those that you incur whether you print one copy or one million. These include typesetting, copyediting, design, permissions fees for quotations and illustrations, drawing interior and cover art, proofreading, making an index, printing ARCs. Printer origination, including file manipulation, platemaking and press makeready, might be included here (it is invariable). However it is often just left as part of the printing cost, which makes up the running cost along with paper, binding, jacketing, cartoning, and, if it is included in unit cost rather than overhead as it usually is, the cost of shipping to the warehouse. Once your book is on press you begin incurring running costs — the cost of presswork, paper, binding etc.

Running costs

Running costs when looked at per copy are relatively small — little quarters and dimes running along to the printer’s bank account — whereas that wad of Benjamins representing the fixed costs always looks pretty significant. The economics of book manufacturing mandate that the longer the machines run the lower the cost per copy, and it is this mechanical and marketing quirk which tempts publishers to run more and more copies in order to bring their unit costs down. This is of course a mathematical illusion. If you run a million copies your running cost will indeed be less per copy than if you print a thousand — but it doesn’t take an Einstein to recognize that you’ll have spent more money. If you never sell those 999,000 additional copies — problem!

The really important numbers in book publishing are

  1. the ideal retail price and
  2. the number of copies you can expect to sell.

The unit cost — the cost of making each single copy, the running cost added to the fixed costs divided by print run — is however the number which has acquired an overwhelming importance in the minds of book publishers. I personally think this is because the really important numbers are very difficult to calculate (maybe impossible to calculate), so publishers indulge in unit cost manipulations as a sort of displacement activity in order to make themselves look like they are being rational and scientific about things — thus disguising the fact that they are ultimately just going on hunches about the really important numbers.

Of course the amount you pay to manufacture your books is an important factor in your business. This is best taken care of by a forceful negotiation with your suppliers over a contract covering an extended period. If your price picture is X, doing endless estimates involving different quantities just wastes time as the answer is always going to be a function of X. Of course, the more copies you print, the lower your unit cost will be. This means that doing endless unit cost calculations, solving for different unit costs will cause unit cost to look vitally important. However, apart from wasting a lot of time, these calculations can only have one result, which is to make you decide to print more and more copies. We all know that printing more copies than you can sell is the high road to Queer Street: tie all of your capital up in unsold inventory, and insolvency is just around the corner. Excessive costing is dangerous to corporate health.

In the U.S. book business when we buy paper we talk about basis weight in referring to text stock but when we move to cover stock we talk about caliper. Caliper is the thickness of a single leaf of paper. Thus a cover measured by a micrometer at 0.010 inches is referred to as 10 point cover stock (or 10pt) — a common caliper for paperback covers. We indicate basis weight by the pound sign: #. Basis weight is always the weight in pounds of a ream of paper of a standard, basis, size. Superficially pretty straightforward.

Cover and text stocks each have both caliper and basis weight, yet we never think of the basis weight of cover stock, and I found myself last week asserting that they were in fact the same. They are not.

There are two variables in that formula for figuring basis weight: the size of a ream, and the size of the standard sheet. The ream is (almost) always 500 sheets, but there are different standard sheet sizes for different types of paper.


What is the reason for these different basic sheet sizes for different print businesses? Tradition/inertia is the simple answer. When paper was made by hand, the sheet size was determined by the dimensions of the mould used by the papermaker. For example, because sheets measuring 17″ x 22″ leant themselves to cutting into four 8 ½” x 11″ sheets, this became the standard for business stationery as the papermakers serving that business made their moulds that size. Strangely we in America have never changed our units from this quaint basis.

Why do we need to know the different basis weight calculation of cover and text stock? It comes to the fore when you venture overseas for your book production. The rest of the world tends to deal in gsm (grams per square meter, sometimes written g/m2 or g/m²) a much simpler and logical, if less picturesque, system of measurement. A book manufacturer in China may well quote you cover stock referred to in basis weight terms which they’ve converted from their gsm data. (To convert from text basis weight to gsm divide the basis weight by 0.67565. For cover stock the magic number is 0.36989. You’ll no doubt need to do a little rounding to make the answer look right.) So be aware: 100# cover stock is not the same as 100# text stock. Because the area of the basis sheet of cover stock is 1806sq.ft. while the area of the basis sheet size of text stock is 3299sq.ft., your 100# cover stock would be more like a 180# text sheet — if such a thing were available. Verso has quite a useful conversion tool.

For caliper calculation, see PPI. See also Basis weight calculation.

Jeff Peachey, at his post about this Patent Office model book trimmer, tells us that it operated by gravity — the blade is really heavy! I imagine that before the development of this sort of cutter book edges, in so far as they were trimmed at all, would be cut by a sharp knife. But of course early book edges tended to remain uncut through the binding process.

Here below is an illustration of the original gravity-operated trimmer:

Marie Antoinette offered an apology to her executioner: she’d stepped on his toe.

Mr Peachey also shows us a photo of the earliest book he has found with evidence that its edges were trimmed by a guillotine. (Follow that link for a pretty gruesome picture!) The evidence for the guillotine-trimmed book dates from 1834 and is there because of a blunted blade. 1834 is the date given (with a question mark) by the Patent Office for the date of the patent submission for the trimmer shown above. (The Patent Office lost most of its records and all models, which used to be required as part of a patent submission, in a fire in 1836, and has recently been reconstruction the models.)

Well, by now you all know that Penguin Random House, or rather Bertelsmann, is going to be taking over Simon & Schuster. No doubt our government will have to think about this idea, so the deal will take a while to finalize. The merger looks like it will take a similar form to that between Penguin and Random House themselves in 2012-13, where the two businesses continued to publish separately for the first few years, until they “grew” closer and closer by 2019.

Bertelsmann’s acquisition activity is wide-ranging. As Publishers Weekly reported on November 3rd, “Bertelsmann, the German media conglomerate that owns the U.S’s largest publisher, Penguin Random House, has acquired Quad/Graphics’s two remaining book manufacturing facilities. In a deal that closed on October 31, Bertelsmann Printing Group USA’s Berryville Graphics division acquired Quad’s book printing plants in Fairfield, Pa., and Martinsburg, W.Va.

The sale completes Quad’s exit from book manufacturing. In July, it sold its other book manufacturing plant in Versailles, Ky., to CJK Group, Inc. Quad said it will use the proceeds from the sale to pay down debt.

In addition to Berryville Graphics, Bertelsmann Printing Group USA owns Coral Graphics, Dynamic Graphics, Offset Paperback Manufacturing, and Digital Print Services. Prior to the Quad acquisition, Bertelsmann Printing had annual revenue of over $230 million and employed more than 1,300 people.

Bertelsman has long been one of the firmest corporate believers in the future of book publishing, including in print books. The move comes during a tough time for the printing industry in the U.S., which was beleaguered by economic issues even before the pandemic.”

The printing acquisition might look more significant than I suspect it really is. As the press release notes, Bertelsmann, via its print division, Berryville Graphics, already owns considerable manufacturing capacity in the USA. Now it owns more. It’s not like these plants only work for PRH: you can’t really base a book manufacturing business on a single client. Of course now that that client is growing even more that might be becoming something which could actually be conceivable!

I’ve spent much of my life arguing that publishing is a business that wants to be small. Those guys in Gütersloh seem determined to prove me wrong. The combined PRH + S&S would be the publisher of ⅓ of the trade books published in America, and this certainly looks like concentration. Maybe the Justice Department will prevent the consummation of this marriage, but I have to believe that as companies coalesce, so bits and pieces (people mainly) get thrown out. These bits and pieces, not being ready for extinction, often set up little publishing operations of their own. This is something which technology has made easier and easier — or at least to which it has lowered the financial barriers — and whack-a-mole-like I’m popping up again saying “Look. This consolidation means more diversity.”

Bertelsmann, “one of the firmest corporate believers in the future of book publishing”, one might hypothesize, could be aiming to produce a counter-weight to Amazon’s heft on the retail side of the book business scales. Maybe. But I suspect that the primary motivation is that ever-threatening capitalist diktat that in order to be successful all companies have to grow and keep growing. I’m not sure monopoly concerns would be a problem in this case. Each individual book is an individual book: the good of the public is served by having them all made available, and it doesn’t really matter to anyone which publisher the book is issued by. If every book was published by one publisher, wouldn’t that publisher still be trying to entice more readers to buy each one of their products?

Of more concern perhaps is the position of authors and their agents. Less competition between publishers will probably lead to lower advances, and perhaps less generous royalties. My “defense” about this, admittedly a little long-term, is that those discarded “bits and pieces” and the new companies they establish will eventually reestablish the competitive environment. PRH-SS can surely not afford for it to become known throughout the literary agent population that there’s no point in offering your book to them: just offer it first to HarperCollins, Macmillan or whomever.

It’s great that Bertelsmann is rooting for our industry. If PRH-SS becomes a reality, the industry will change, but perhaps not as much as one might think. And, be it said, if PRH-SS does not happen, then too our industry will change.

Apparently the Guru Granth Sahib, the holy book of Sikhism, is always printed as 1430 pages. This is very unusual for a book: if you typeset a text twice you are quite likely to come up with a different page count: maybe one or two pages different if you are using the same design specs, but if you change the type size and page format obviously potentially vast — just look at those Bibles lying around the house. The pages of the Guru Granth Sahib, called angs, include 5894 shabads, hymns and prayers, and can be divided into 60 rāgas. The book is the focal point in any gurdwara. The reader/chanter performs from a raised platform known as a Takht (throne).

Clearly having the book make fewer pages would make handling it a bit easier.

Official versions of the Guru Granth Sahib are only allowed be printed in the basement the Gurudwara Ramsar in Amritsar by the Shiromani Gurdwara Parbandhak Committee. Obviously this makes it simpler to control the 1430pp page count. As the book is composed in verse lines, keeping the same number of lines per page might have been a fairly obvious idea to come up with back in 1704 when the text was finalized. That page count has nothing to do with even working — why would any religious directive be divisible by 16 any way? Actually 1430 isn’t even divisible by 4, but of course maybe there’s also some front matter to help the mathematics a bit more.

Early 19th-century manuscript ang (Schoyen Collection Norway)

The Guru Granth Sahib was largely composed by six Sikh gurus: Guru Nanak, Guru Angad, Guru Amar Das, Guru Ram Das, Guru Arjan, Guru Teg Bahadur and Guru Gobind Singh Ji. It also contains the poetic teachings of thirteen Hindu Bhakti sant poets and two Sufi Muslim poets. The script used for the text is Gurmukhi, an abugida developed from the Laṇḍā scripts, and is available in several languages including Lahnda, Braj Bhasha, Kauravi, Sanskrit, Sindhi, and Persian.

Dr Alex Bubb, who noted at the SHARP listserv the unusual fact that Guru Granth Sahib has a constant page layout and length, asked if there might be any example of such a phenomenon. Dr Paul Tankard has replied noting one — “G. B. Hill’s Clarendon edition (1887) of Boswell’s Life of Johnson, in six volumes, quickly became regarded as ‘standard’. When in 1923 L. F. Powell was invited to revise the edition, it was stipulated that (with regard to Boswell’s text) Hill’s pagination was to be preserved. This necessitated the creation of an extra level of annotation, and the addition of various appendices to each volume. So, 133 years later, Hill’s pagination is still being used for a text that remains in print.” But of course you can get editions of Boswell with different pagination.