I flip between delight and frustration when I read Mike Shatzkin lauding Lightning Source and the POD (Print on Demand) concept. Delight, because he’s right; frustration because I’ve been singing this song since 1993 when I left publishing for a couple years to work for a POD pioneer.

Mr Shatzkin maintains that corona virus has transformed publishers’ thinking about inventory. I hope he’s right, but I suspect publisher minds are a little harder to change than a couple of months’ isolation will cause. He assures us that a book can be done overnight: Publishers he says “can send down the book file today and have Ingram start shipping whatever quantity they need to the stores, or Internet resellers, tomorrow.” This is a slight exaggeration: it does take a bit of time to set up a book in the Lightning Source system, and a careful publisher might like to see a proof before “turning the title on” for supply to customers: but once the book is in the system, most orders (if they are for paperbacks) can indeed be printed immediately, and shipped to the ultimate purchaser next day.

What he says about unit costs is absolutely right. “Print more and you get a lower unit cost but you tie up capital for a longer time and take the risk that you’ll fail to sell them all. Print fewer and you save cash and reduce the risk of waste, but you add to the risk that you’ll be caught short at some point.” Along with the ability to judge which books to sign up, this balancing act is the central skill of a publisher. Publishers regard themselves, with a certain justification, as really good at forecasting demand for their books. The forecast sales number determines print number which determines unit cost. Be a little cautious in the face of claims that intuition, experience, and innate judgement play a role — demand planning relies on historical data: how many did this book sell last year; how many did this comparable book sell; how did the author’s last book do; are there any events coming up which might cause a bump in the sales curve? In a company with thousands of titles this sort of calculation is done hundreds or even thousands of times every week. Hey, if we can’t do it, nobody can do it! If we weren’t good at this we wouldn’t still be in business.

Quite apart from the fact that all historical sales patterns have just been blown out of the window, the print numbers for academic books are now just so low that spending time thinking about the right number to print is a waste of time and energy. The right number to print for a book that’s going to sell 200 copies is zero! It should be set up for print on demand right away. The unit cost — which is the number all publishers focus on like a Siren call — is likely to be insignificantly different as between 200 and 250 copies, so that there’s really no payback on the time spent figuring out which number is better.* And the difference between the unit costs for 200 printed by offset and for 1 copy printed by POD is not so large that the difference is going to make or mar anyone’s annual performance. BUT, even more significantly, if you used POD for all your books and consequently didn’t have a warehouse you might be saving (I don’t know, and it will vary from case to case anyway, but let’s say) 25% of your overhead costs. Plus, you will no longer have most of your capital tied up in an inventory of unsold books sitting in that warehouse for however long it takes to get rid of them. Plus, you will never be out of stock. However good your demand planning, there cannot but be cases when you sell more than you expect to and end up without stock, which necessitates a rush reprint which may mean you can’t sell any copies for a couple of weeks. Plus, there are always a few mistakes where too many books are printed, sit around accruing storage costs for a couple of years, and then have to be wasted. None of that needs to be thought about if you’ve set up all your books for print on demand.

Mr Shatzkin implies that Lightning Source is the only game in town, but of course this isn’t the case. Lightning Source is a division of Ingram, and this provides a large measure of efficiency as books shuttle from binding line to wholesale shipping. They are a wonderful operation and all publishers should be using them, but other suppliers, notably Bridgeport National Bindery, can provide a similar service. In fact when it comes to hardback books (and yes, you can do a hardback book via print on demand) Bridgeport will make you a superior product.

Anyone setting up an academic press these days — and such daring souls do seem to exist — should forget about warehousing and inventory and exclusively use print on demand. When I was first involved in the POD game, the main concern from publishers was the quality. This is no longer an issue, though some traditionalists love to hang onto it as an excuse for inaction. The reason early POD books didn’t look good is that back then nobody had text files: computers were just beginning to be used in book production then. Early POD books were made from a scan of a printed book. Some of these printed books were less than beautiful printing jobs (yes, traditionalists, even an offset book can be badly printed) and the resulting POD version could never be better than the book it scanned. Be it also admitted that scanning was often less than brilliant. Sure every publisher dreams of the big book where they could clean up on a really low unit cost — but beware, hubris beckons. Stay with the basics, and if you do end up selling thousands of one of your books at a unit cost higher than it might have been, at least you have made money, less perhaps than you might, but losing money is always simpler than making it, so be grateful for what you’ve got in the bank.


* Just about the worst decision a publisher can make is to print more copies of a book because the unit cost is lower (thus the potential profit margin is higher) at the bigger number. Print what you know you can sell. Not one copy more.