Publishers of technical software books and magazines seem to struggling against the relentless onslaught of the Internet, crushed between the twin rocks of rapidly changing technology and free online content. In a recent .NET Rocks! podcast, accomplished technical author Charles Petzold (of Programming Windows fame) discusses the grim commercial realities of writing technical software books in the 21st century. It doesn’t sound good. His recent 3D programming for Windows book took 8 months to write and has sold less than 4,000 copies worldwide. As he gets royalties of around $3 per copy sold (less when sold outside the US), this equates to less than $12,000 for 8 months work. He could have made around $9,000 flipping burgers for minimum wage over the same period[1]. Ouch.
[1] Assuming 40 hours per week.
Successful Software 
The printed book paradigm for technical software books is dead.
Using the numbers above and a sale price of $19.00 instead of the list price of $39.99. 4,000 x $19 = $76,000. That is a much more acceptable income level for 8 months worth of work.
“this equates to less than $12,000 for 8 months work. He could have made around $9,000 flipping burgers for minimum wage over the same period[1]. Ouch.”
Then go flip some burger, that is, if you can actually enjoy it better than writing stuff.
The realities of book publishing is two fold : books that take old paradigms and fundamentals such as this one and bring them to a new API shouldn’t exist at all. Because if programming 3d software is your day job you should have learned ALL the fundamentals when you were studying at the university. And learning new libraries and implementations should be done just by looking at the actual API documentation, if you are not able to do so you are not fit to be a programmer, endgame. That was the way we used to distinguish the skilled programmer from the crap during the Java dot com bubble : those who could learn Java just by reading source code and the official API documentation were skilled, the ones who needed books were crap.
Does this have nothing to do with the topic? Is is possible that if the book he wrote was “COBOL for Web 2.0”, he would have done even worse financially? This is a free market here. I am sure the guy who wrote whatever is the top computer book on Amazon.com right now doesn’t have the same issues with the amount of time he/she spent on it. It really boils down to the basic fact of if you want to make money writing, you have to write something people will buy.
M@
WPF is prrety hot topic:
http://www.google.com/trends?q=wpf
I don’t know how many people are interested in 3d programming for WPF though.
This sounds like a similar situation to the “bands make most of their money from touring, rather than selling CDs” idea. The value from the direct sales of the book may not be all that great when looked at on their own, but the value available to you as the guy who wrote the book has to be a factor. Whether it’s speaking arrangements, consulting opportunities, advances on future books, whatever. Surely looking just at the simple book sales ignores the whole picture.
Still doesn’t make it not suck though ;).
And yeah, I think you have to consider the subject matter, some computer books must sell a lot more than that. You make a decision when you choose a particular niche or target, it’s the same with software or any venture. It’s always part gamble, there’s always something else you could have done that might have been a safer choice, you’re always partly motivated by a slightly selfish personal satisfaction thing (even if flipping burgers paid 50x times that, some people would still be writing computer books).
Let’s face it though – “better” books sell more too and generate a bit of surplus income as well. It is a lot of competition, but some actually enjoy writing books.
Maybe not every book simply has a good target audience. I remember years ago I bought 3 php books. I read only one of these.
Today I no longer have ANY php book and I dont regret it at all (i converted to ruby 5 years ago and never regretted this move from php to ruby). I used to have perl books and gave all but one away (that one was special because it was bioinformatics + perl, I still digest the information inside, but once i have read it and noted what was interesting, I will give it away as well)
What should matter is the knowledge. As people learn more their knowledge horizon will expand and change, and authors need to adapt to this in some way or another.
And let me assure you – people LIKE to have GOOD knowledge! So in principle there will *always* be a solid audience.
It may just be that their view and goals differ from an author’s goal.
The writer gets next to nothing but the publisher and the outlets must have made a neat profit even at just 4,000 copies sold.
Per book:
List price:$39.99
Author: $3.00
Bookshop: $20.00
Publisher: $16.99
And at 4,000
List price:$159,960
Author: $12,000
Bookshop: $80,000
Publisher: $67,960
Of course this is oversimplified but still…
Per book:
List price:$39.99
Author: $3.00
Bookshop: $20.00
Publisher: $16.99
And at 4,000
List price:$159,960
Author: $12,000
Bookshop: $80,000
Publisher: $67,960
You nerds forget the costs which a publisher and a retailer has.
A book written by author needs to be vetted by a Technical reviews, a language editor, and also in many cases there multiple levels of editing. In addition the current printing costs per book will be around $ 6.00- $ 10.00 depending on the volumes, quality of printing and number of pages. A publisher needs to take care of costs of also funding both acquisition costs as well as production costs.
A retailer pays huge establishment costs again for reaching any book to the market. Remember there are distributors in between the bookshop and the publisher. Few publishers can afford a team that can reach the book directly to the vast number of retailers, they go through a middle tier channel
Wow, $3 per book? You’re lucky. The last book I wrote I got a whopping $1 per copy sold, and that was split between two authors. You will never get rich selling books for $.50 a copy. You do it for the name recognition. I’d say eight months to write a tech book is about average with all the review and editing time. I don’t think many people have much of a clue about the book/publishing business and how much authors actually make. They just see the $59.99 price on the back and think that we make a ton of dough. It just doesn’t work that way.
Bruce Schneier, D.J. Bernstein and Peter Seibel are good examples of where the comments are going here. Years ago you could keep a tome of a book around as reference material but today whats the point when large portions go outdated and you can get up to speed on certain details via Wikipedia, the authors website, or elsewhere.
maybe there’s a way to automate this or even automate the localization/internationalization?
http://ejohn.org/blog/programming-book-profits/
There seems to be a bifurcation in the market. Pragmatic, Manning, and Oreilly issue Beta books, sometimes when only 2/3 of the chapters’ first drafts are written, and people go at it. Or you have Real World Haskell, or Manning, Raghavan , Schütze, Introduction to Information Retrieval, Hickey’s OCaml book, they put up drafts and people start tearing into them, adn the final product can be a committee effort (i.e. thoroughly proofread, no major topics missed)
On the low end, packt, Apress and wrox put together readable books without spending enormous amounts on graphic design and typesetting. They’re not beautiful, but they’re usually cost-effective.
Should have sold his book through Amazon Kindle. No one likes to carry around technical books that weigh 30 pounds. And through digital distribution, there are no printing, storage, or shipping costs.
I was just in the large bookseller today looking at computer books. The store has cut the shelf space for computer books in half at least, and what they have on the shelf is not organized. I would guess that at least 1/3 of the books were for dummies (or their clones).
I wish that there was somewhere in town that carried a better selection of books. I would much rather have a book that I can carry around the house than try to wade through all the crap on the net.
IMHO I believe we will start to see more and more of P.O.D. in the future and less of the whole write-review-publish.by.publisher scenario thats dominated in the past.
This post confirms it for me.
Maybe the author could have tried some new marketing venue such as a blog to promote a book he has available in lulu.com (although we’d have to see how many pages the book consists of). He’d still retain the whole ‘expert’ status and still be able to make a good buck for it.
In my experience, I’ve noted I can gauge how much of an expert someone is in a subject when they run a pretty compelling blog on the matter. As a matter of fact, nowadays I trust many blog authors more than I do many of my personal library’s book authors.
@Gene Tani: Indeed. Some authors are going the open source way, which I find particularly interesting. The challenge seems to be on how to establish a good set of parameters on writing and reviewing the drafts to make a good compelling book.
On a final note, concerning tech publishing, I believe the printed book is a doomed medium. I envision a tech book in the future with more sophisticated e-ink e-book viewers with auto-updating entries. “This book has been updated by it’s author. Do you wish to update?” “Do you wish to send a comment, or typo to the author?” Nice. :)
This is really a sad fact. Charles Programming windows was a Bible for Win32 programming. I have no doubt this would be great too, but I haven’t read it yet.
However, I think WPF is merely starting to pickup in years to come, it will be fairly lucrative for him (I hope!). The guy who said above that books are for dummies or not real is highly mistaken! Have you did a 3D programming before? You can’t learn 3D Programming by reading documentation only.(You can but it will take you ages! and when you are done grasping that version is out of date!).
Productivity is required even in learning new things because of the constant changes in the industry.
I read a lot of books, may be not from cover to cover but they get me started for most interesting things. And I hate to read PDFs. It is just not joyful to read on screen as compared to real paper. May be if we have good eReader technology Then publishers could cut out the Distributor chain all-together making it more lucrative for the authors.
@Parag
>The guy who said above that books are for dummies
I think he was referring to the “For dummies” series of books, rather than insulting the readers.
I like having a book in my hand. I haven’t tried the Kindle before, but I have over 200 books in my library. If I had the Kindle I would want to print out some pages here and there to send them to another programming that I’m trying to talk to.
I still buy books because the writing on the web is not always organized and coherent like a book is after the technical editors and grammatical editors get through with it.
Print on Demand, as one person said is probably the wave of the future.
Also there are no good walk-in bookstores that have tech books anymore. Microcenter used to, but I went in there a month ago and they reverted to the consumerism books.
Amazon is the only place to peruse the tech books. I bet I buy around $500 – $1000 per year (maybe 1 – 2 a month) from Amazon.
I’m not a fanboy for Amazon, but I know if I go there I can find it and usually read some info about the book and their “look inside” feature is very nice.
Hi Andy,
My Comment was more related to : Z :
“And learning new libraries and implementations should be done just by looking at the actual API documentation, if you are not able to do so you are not fit to be a programmer, endgame”
Clearly that’s not true. If I have to learn 3D Programming, a book is needed to get me on fastTrack. It is about productivity in learning then programming ego. Sorry If I offended anyone!
But he previously sold a ton of “Programming Windows” in the early and mid 90s especially, right?
So what’s the difference, apart from the date?
I think the trouble is that he picked a subject that doesn’t have a broad enough appeal (3D programming is a specialist area, and this book is a specialism within it), which is over populated with competition (there are really tons of books on 3D, and which I’d imagine that much of the potential readership feels is unnecessary (see comments in this thread).
I’m not unsympathetic but maybe even a name author like Petzold can’t make a living doing quality books with so many things against him. It’s the same problem as an ISV who spend a year writing a fantastic (but highly specialized) program that only a few people would actually pay $50 for.
Maybe this is why e-books get such a bad rap – intentional marketing on the part of the book companies. I would advice future book sellers to be like a mISV – sell it via digital download (make the first chapter free)..and you can optionally order a hard copy for $10 more.
I love printed books, I find PDFs hard to flick back and forward through etc, and books are nicer to read on the train, on the couch or in bed.
I buy books regularly and have over 50 books on my Amazon wish list.
However I have noticed a change in my book buying habits, after noticing how little I use books like “undocumented dos” and “borland c++ builder unleashed” while books from the same era like “object oriented software construction” and “introduction to computer graphics” still get flicked through now and then.
I never buy product specific books now. Almost all books I buy now have to be about principles and patterns that I can use with a variety of technologies, languages and tools now and in the future.
I would never buy a book called “3d programming for {product x}”. I would buy a book called “fundamentals of 3d programming” and simply refer to the {product x} documentation for details of how that product can implement the fundamentals, and if that product requires me to forget all those standard, portable, technology and tool independent fundamentals, then its not the product for me, I will find a product that maximizes the return on portable knowledge, and minimizes the product specific overhead required.
(Sadly in the windows world, it seems a lot of products seem to bring their own “fundamentals” that can only be used once with that product and then must be forgotten again, once the product is obsolete, otherwise one wouldn’t need a book like “3d programming for windows”.)
I mean 3d programming is old and well understood by now, if WPF really does things so differently from everything else as to require a whole nother book (rather than some simple online api docs to show how existing knowledge can be be directly applied), then its no wonder it hasn’t caught on, which could be another reason why such a book hasn’t sold that well.