May 20, 2012

Print-On-Demand Not Coming to a Store Near You

A recent article by Andrew Fox about the possible future of publishing and bookselling made me start thinking about some of the Print-On-Demand printing machines that are available for bookstores.  One of Andrew's contentions was that in the future bookstores would commonly have equipment of that sort in their shops.  It was one of the only things he said that I didn't agree with, primarily because of the price of such machines (a common one, the Espresso Book Machine, costs around $100,000).  Yesterday I realized that my opinion about that, since the last time I'd seriously considered it was 2008 or so, might have been based on old information and an old thought process.  For a small business in a changing field, sticking to conclusions that you made four years ago is a terrible habit so I did a big of digging around the web to check my assumptions.

I wish what I had found would have made me wrong and Andrew right, but not so.  For a smaller store, the technology just isn't there yet.  And, I'm very doubtful that it ever will get there.  You'll find my reasoning after the break.

Before I go into detail: one disclaimer, my figures for price of both the machine and supplies is based on the best information I could find.  However, it comes from an article at Wired from 2009 and a Q&A from the University of Michigan Library dated late 2010.  To try to improve my information, I contacted the manufacturer of the Espresso Book Machine (EBM hereafter) to get purchase and lease information.  When I hear back from them, I'll post what I find out.

To start with, here are some basic numbers for printing a public domain book from Google -
Cost of machine is $100,000.
Materials for a 300 page book cost $3 per copy.
Google and hosting service charge $2 (books from other publishers would cost more but would sell for more.  Net result would be a wash).
Suggested retail is $8.
Finally, the EBM can print one book in about 4 minutes.

Please feel free to ignore the following math and jump down to the conclusion.  $8 retail minus $3 in materials and $2 in fees leaves $3 profit per book.  Borderlands is open 8 hours per day, 362 days per year.  At the rate of one book per hour, the profit in a year is (8 x 3 x 362) $8688.  Based on the cost of the machine, I would pay off the machine and start making a profit in 11 year and 6 months.  At the rate of 3 books per hour, the machine pays off in 3 years and 10 months.

At the rate of one book per hour, every single hour we're open, the machine would pay off in 11 years and 9 months.  At the rate of 3 books per hour, it'll take 3 years and 11 months.

I'm not sure what sort of sales volume would be reasonable to expect if we got an EBM, but my bet is that it would be somewhere between those two figures. And that's the problem -- a hundred grand it far too much money for me to tie up for year and years.  And, of course, none of the above figures takes into account the interest that I would have to pay on the loan, because I, like most small and mid-sized booksellers, don't have that sort of money in the bank.

Sadly, the math still works out the way it did in 2008.  Too much money tied up for too long a time equals too much risk.  If the price of the machine were lower, the equation would change, perhaps dramatically.

Problem is that I don't think that's going to happen since the actual physical technology used in machines of that type is really quite mature -- in essence it's laser-printing combined with basic 20th century robotics.  The thing about mature tech is that the price doesn't tend to change much unless the market for the specific implementation of the technology increases significantly (in that case, economies of scale kick in and the price drops).  I don't foresee the market for "book machines" increasing much, especially in the face of increasing adoption of ebooks.  Gadgets like the Espresso Book Machine are probably going to remain where they are now: out of the reach of most bookstores.

Comments? Anyone out there see any way around the math?  I'd love to be able to have a machine like that in the store.


  1. Why not sharing the machine? If 10 local bookstores buy it together, the maths are becoming more friendly... Also, this is a huge educational market where teachers sell their books to their students for more than $100 (Law is a good exemple). These authors make very little on each copy. Why not POD that segment?

  2. Hi Luc,

    Good point. But that shared model wouldn't quite have the same benefits for the other 9 stores that didn't have the machine at their location. A plan like that would certainly produce income for all the stores participating, but the over all math would still be pretty much the same since the income would still have to pay off the machine for each of the stores to make a profit.

    Thanks for the comment,

  3. A couple of things that might make this work:

    1. If it can print a books in 4 minutes, then you can print 360 books/day. 3/ hour for eight hours is only 7% of capacity.

    The reason I mention this is that you aren't in any way limited to JUST printing books for customers that walk in. Instead of idling, you can also use the machine's downtime to rapidly replenish fast-selling inventory, do print jobs for local businesses (yellow pages, churches), operate a small publishing company on the side etc.

    2. I've heard the biggest users of these machines, where they've been installed, are actually self-pub authors- It's an easy way to do a small print run if you don't have a publisher. So if you arrange a couple book signings, etc. you could easily have days where you print dozens or hundreds of copies at once.

    3. The $2 fee paid to Espresso and Google is for accessing their book catalog - I think you can also feed the machine a PDF for free.

    So you can print classics from Project Gutenberg, your own computers etc. for just the $3 cost of material. And if you have any direct deals with self-pub authors that would be $3 to print as well.

    That should change the calculus a little, as $3 is often less than the cost of shipping a single title.


    Some cons

    1. It's a lot of work. You or an employee have to babysit the machine.

    2. The machine IS still expensive, and it may come down in price soon.

    Barnes and Noble is the elephant in the room and they haven't yet made a move in this space, but it's hard to imagine they aren't planning something.

    If they decide to partner with Espresso, economies of scale could rapidly drive the cost down. If they decide to compete with Espresso than competition could do the same.

    Either way, it may be prudent to wait if the cost isn't chump change for your business.

  4. Unknown's comment right above raises some good points. And the timing is excellent since I just put up a follow-on post that talks about those very points

    Thanks for commenting,