Evan Marcus is my co-author on the
High Availability book and creator of the rather fun
Bill Gates net worth page.
He knows a few things about statistics, having grown up (and into) a nerdy
baseball fan. I can call Evan nerdy because he and I have co-published,
and Evan has shared some serious insight into the statistics about our book,
which is where this long tale is going.
Technical book sales are in the toilet. Individual technical books just
aren't doing the volume they did a few years ago. During the dot-com
boom, expensing a few hundred dollar online spree at fatbrain.com
wasn't such a big deal. Today, we're lucky if we see a few hundred book sales
per royalty statement, versus a few thousand when the HA book was first released.
The exponential drop-off in sales happened inside of a year. I hear from
tech publishers (the three I talk to more than twice a year) that
book sales stink like old cheese right now. And after watching
NBC's Dateline on supermarket
cleanliness last night, I believe old cheese has a better chance of moving.
So why aren't the real world bookstores devoting less shelf space to
technical books? My gut reaction is that even though individual sales
are down, overall sales are probably about flat due to the larger
number of titles being issued. Total titles are fueled by total topics;
when you work in a fast-changing arena you create opportunity in
the derivative markets.
This is a classic
long tail, popularized by Chris Anderson of
Wired magazine. Even technical "best sellers" slide down
the distribution curve, taking up place under the long tail. The
long tail in technology books is created by the multiplicative
effect of patches, upgrades, new languages, new operating systems,
new architectures and the socialization of technology.
Here's the problem: I think the number of people who are willing to
shell out $30 or $50 for a technical book is going down. Leading
indicator is when there is a thriving used-book business in your
title on amazon.com. It's not
just corporations that frown upon expensing books (don't get
me started on the state of "training", usually the first thing
to disappear when the belt tightens). The long
tail is occupied by books that bobsledded down that slippery slope
from "best seller" to "amazon.com only" very quickly. Why pay for
an entire book when (a) you're probably just looking for an answer
to an immediate question, (b) you can Google faster than using an
index and trying to guess how the publisher's editor categorized
your query and (c) free or cheap beats ten large Dunkin' Donuts ice coffees
worth of content.
What we're going to see is the iTunes-ization of technical content.
Content by the slice, located through search engines, paid for
by subscriptions, advertising, and PayPal. Google and amazon.com
have
hinted at this direction for months. Here's my pricing
prediction: Technical content will go the way of the New York
subway token - priced the same as a slice of pizza. System
administrator fuel of all types should enjoy consistency in pricing.
I believe this long-tail monetization may eventually smother technical
publishing as we know it, and that creates a problem: how many
times do you learn something new by searching for an answer, as
opposed to reading a book and uncovering a question you didn't
previously ask? The challenge for technical publishers is to
find a variety of models that allow authors to be compensated,
make new book projects promising and financially interesting,
and to reach the highly fragmented, appropriate audiences that
represent the long tail.
This doesn't mean that techinical publishing is dead. Far
from it. I think Tim O'Reilly
gets it (Disclaimer: the NFS book is an O'Reilly title
with my name sharing spine space).
It means that the barriers to entry and to exit
need to change. The barrier to entry is obvious -- technical
books have list prices of $25 to $60. Print a book for $10
and you can test the elasticity of demand. Self-publishing
houses like iUniverse do
exactly this, optimizing for making money on a tiny print run
instead of monetizing the first print run at what the market
might bear (Disclaimer again: I've already started talking to iUniverse
about the hockey book, because I think I can self-promote it
better than a classic publisher).
Self-publishing prices to
the cost of production, not market demand, so it's insensitive
to miniscule demand. What you want if you optimize for the long
tail. What amazon.com benefits from. Reducing the barrier to entry
makes technical book publishing accessible to anyone with some
good ideas and a blog. Same model used by emergent
bands that promote themselves entirely through MySpace (props
to Cory Doctorow for the
forward pointer on that).
The barrier to exit is the tricky one, because it requires that
today's "classic" publishers realize that their content is best
suited for search, print on demand, sales by the page or by the chapter,
or simply as free content that drives readers to other
paid content, advertising or promotions. I would love
to be able to put pages of the HA book in blogs, or on
Sun's BigAdmin
site for system administrators, because it would drive awareness
and possibly readership. The "exit" I'm referring to is the
exit from the business model that precludes me from
putting up pages or chapters of the book (the publisher can
do it, and has done it, but typically only as part of an
early phase marketing plan or when they're compensated).
Because our publisher monetized the first print runs of
the book (before it slipped into the long tail), and because
they can now do print on demand and make money from small
print runs, there's no incentive for the publisher to release
the copyright, to effectively open source the book and try out
some dual-licensing models, or to rekindle interest in an
area (making networked systems more reliable) that has
suffered since the dot-com bust made training uncool. So
the content is trapped: can't be re-used, can't be widely
distributed, and can't be monetized except through whole
book sales. Without an exit strategy, it's going to be
smothered, and while the publisher can monetize the long
tail in the aggregate, there's very little incentive for
authors to create more content to fit that model.