Monday, March 7, 2011

Solving the library e-book problem

HarperCollins has announced that it will allow libraries to lend e-books 26 times before demanding those libraries pay, essentially, a replacement fee:
Its sales president, Josh Marwell, believes that's only fair: 26, he claims, is the average number of loans a print book would survive before having to be replaced. ... Clearly, printed books last a lot longer than 26 loans," says Philip Bradley, vice-president of the Chartered Institute of Library and Information Professionals.
I'm also skeptical that a print book only lasts 26 checkouts. And I'm interested in the topic since I finished reading an e-book edition of Neil Sheehan's "A Fiery Peace in a Cold War" that was "borrowed" from the Philadelphia Free Library. What's frustrating about HarperCollins' idea is that it tries to force a print-book scarcity business model upon e-books. That's unnecessary, and probably dumb. Why not come up with a new model that fits the new medium?

Here's what I suggest:

• Instead of forcing libraries to "purchase" e-books and then purchase replacement copies, publishers should set up a subscription-type licensing service. Charge the libraries (say) a $50 annual fee to make 100 e-books available. (I'm throwing out a number, here, for the sake of argument.) That gives the publishers the renewable source of income they need to continue operating without imposing silly rules.

• The libraries would be bound by somewhat similar rules as they are now. If they wanted to have, say, 10 copies of a "Harry Potter" book as part of their 100 licensed books, they could, but each book could only be checked out one at a time: If 10 people were already reading "Harry Potter" then subsequent readers would have to wait until a copy was free. That would prevent people from bypassing paid e-books entirely—lots of people want their copy of a book now, or they want access to their copy in something like perpetuity—leaving libraries in something like the same role they fulfill now. And libraries could revise their stock at any time, discarding five Harry Potter licenses when the book becomes less popular in favor of other selections.

Full-disclosure: This idea is inspired by my Macworld colleague Lex Friedman, who has written about wanting to see a "Netflix for e-books."

In any case, it's clearly silly to make a library purchase a "new" e-book when the "old" e-book hasn't (and can't) degrade in the same way as a print book. Rather than force libraries to live by a model that doesn't fit the digital medium, make a new model. HarperCollins' solution is short-sighted and doesn't actually serve its customers all that well.


KhabaLox said...

I don't like the HC edict for much the same reasons you mention, but I don't think your solution is fundamentally different.

Let's say the cost the publisher charges for the e-book is $10, and you have to rebuy it after 26 "uses". Further assume that this book gets checked out 26 times per year. This book then costs the library $10 per year.

Now let's look at your solution, and say that the license fee is $10 per year per book. The bottom line ends up being the same.

The only real difference is that under your system, unpopular books ride on the coattails of the Harry Potter's of the world. But I would imagine that if such a subscription service were implemented by publishers, you would see differential pricing by title ($25 for the Harry Potter license, $15 for The Road, etc.)

The HC idea actually has the benefit of being somewhat easier to implement from a pricing perspective.

That said, "Netflix-for-e-books" is a pretty cool idea. If I read more, and read on an e-reader, I'd be interested in such a service. does a hybrid idea, where you pay a monthly fee which gives you a credit or two a month, which you use to buy the audible books.

Joel said...

K- I actually thought a little bit more about that after the post, and you may be right. I think I wouldn't mind an additional layer of pricing in the mix-a nominal replacement fee when the library swaps out books, probably based on popularity. Details can be worked out, but the current model is not very smart for a variety of reasons.