Friday, 17 April 2009

What O'Reilly don't want authors to know

Mike Hendrickson has been busy updating O'Reilly's analysis of the state of the computer book market by programming language. That means it is time for us to reiterate how authors of decent books can earn far more for their work by cutting out the middlemen including trade publishers like O'Reilly.

Traditional book publishers are a dying breed. Aside from e-books, they have been driven out by an increasing number of so-called "self-published" books. In the context of software development, this is particularly common around non-mainstream subjects and includes titles such as OCaml for Scientists and Programming in Scala. O'Reilly's analysis excluded all such books even though they are far more profitable for authors.

In order to make a case for self-publishing it is necessary to present some information about a variety of existing books:

As the OCaml and F# markets are similar sizes and the contents of OCaml for Scientists and F# for Scientists are similar, it is interesting to note that OCaml for Scientists was 22× more profitable for its author in Q4 2008 even though F# for Scientists was a new book published through a major publisher and advertised and sold on Amazon. The reason is that John Wiley & Sons have done virtually nothing to market this F# book besides placing it on Amazon and, in fact, Wiley only managed to sell one copy direct outside the US! This really highlights just how little value a trade publisher and Amazon add.

Haskell is currently receiving far more press than any other functional programming language and, consequently, Haskell books are among the best sellers with 1,491 copies sold in 2008. However, with prices like that of Real World Haskell, they would need to be selling 20,000 copies a year to compete with OCaml for Scientists in terms of the profit made by the author.

Despite being self-published, Programming in Scala is outselling Real World Haskell according to its Amazon sales rank.

Finally, companies like IdTechEx prove that self-publishing can be the foundation of a serious business on the scale of a major trade publisher like O'Reilly. This is a consequence of the characteristic V-curve of profit found in many markets where cheap and expensive luxury products give locally maximal profits. Self-publishing naturally lends itself to higher quality and higher value books.

The moral of this story is that self-publishing is the most rewarding way for good authors to reach their readers. If you are the author of a decent book, please consider publishing it yourself.


7 comments:

Michael said...

I think you are underestimating the services a good editor provides for an author, and perhaps more imporantly, the reader.

Personally, I am much more likely to purchase a book that has been through the editing process by a good publisher than a self-published tome.

Flying Frog Consultancy Ltd. said...

@ Michael

I agree that the editorial process is important but, again, self-publishers do a more professional job. For example, OCaml for Scientists was edited by Xavier Leroy, the lead dev of OCaml, and he provided many valuable criticisms that improved the book. In contrast, F# for Scientists was handled by a nameless editor at Wiley and they provided no valuable feedback at all. Indeed, they could not because they have no domain specific knowledge at all. I had to get Don Syme, the creator of F#, to act as an editor myself and I even had to offer to recompense him myself although he politely declined.

Tim O'Reilly said...

I'm curious how you calculate profitability, given only the price. Profitability is price times units sold minus costs.

Now, it's certainly true that a direct sale of a book at an £85 price could be very profitable - but that's, quite frankly, an outrageous price. We could make a lot more money selling our books for £85 too. But the readers wouldn't be very happy.

The other thing to keep in mind is that while it's easy to count as costs only the cost of each book you sell, the self-published author's costs are actually far higher. Many first time publishers make the big mistake of printing too many copies, far more than they end up selling, because they get a lower unit cost with the volume. (Obviously, you don't have this problem with a print-on-demand service like Lulu, but then the unit cost is quite high, and they take a cut on top of that.)

Imagine for a moment that you print 1000 copies, and sell only 300. Your manufacturing cost is actually triple what you told yourself it was.

A self-published author also has the additional time and cost of shipping copies. Have a big success, and you might wish you had a publisher to handle some of those tasks for you.

I do agree that self-publishing can be an important option to evaluate. After all, I began O'Reilly as a self-publisher. Over time, we learned the ropes and began publishing books we didn't write ourselves. But I can tell you that over time, we sold a lot more copies (even though we incurred a lot more costs) by expanding our reach through retailers.

Also, FWIW, we've sold about 6000 copies of the Haskell book so far, not the 1491 copies you mention. Bookscan only includes about half the retail market. In addition, we sell many copies from our own web site (which are not reported, as well as electronic access through Safari Books Online.

fpmatters said...

Note that Programming Scala (about to come out) is a different book than Programming in Scala (published). Scala is really taking off with lots of books written about it.

Flying Frog Consultancy Ltd. said...

"I'm curious how you calculate profitability, given only the price. Profitability is price times units sold minus costs." - Tim O'Reilly

I was referring to the profit for the author, which is revenue minus costs for self-publishing or royalties for trade published work. As I said, self-publishing is far more lucrative if you make a good product.

"We could make a lot more money selling our books for £85 too. But the readers wouldn't be very happy." - Tim O'Reilly

If the reviews of Real World Haskell are anything to go by then most of your customers are not happy despite your low prices:

"I came away from Real World Haskell still not able to make heads or tails of the language." "failed my expectations" "I must say I'm disappointed" "I was quite surprised and annoyed" "the organisation is terrible" "I am baffled by the amount of praise the book is getting" "this is exactly how not to do software development" "i have a decent prior knowledge of functional programming and even i found myself questioning the presentation of topics" "If RWH gets such great reviews it's because the rest sucks" "It was rushed out to cash in on the Haskell fad which is only among Haskell fanboys"

I appreciate that the reviews on Amazon in the US are all extremely positive but other reviews that were not rigged are generally negative.

"Imagine for a moment that you print 1000 copies, and sell only 300. Your manufacturing cost is actually triple what you told yourself it was." - Tim O'Reilly

That's why we print on-demand. We have no such wastage.

"But I can tell you that over time, we sold a lot more copies (even though we incurred a lot more costs) by expanding our reach through retailers." - Tim O'Reilly

Sure. I don't doubt that you've shipped a lot of copies. My point was solely regarding recompense for good authors.

"Also, FWIW, we've sold about 6000 copies of the Haskell book so far, not the 1491 copies you mention. Bookscan only includes about half the retail market. In addition, we sell many copies from our own web site (which are not reported, as well as electronic access through Safari Books Online." - Tim O'Reilly

That's great but I doubt any of your books will ever reach the success of OCaml for Scientists for the author.

admin said...

nice post

jay paul said...

Excellent post and wonderful blog, I really like this type of interesting articles keep it u.

HP - ENVY TouchSmart Sleekbook 15.6" Touch-Screen Laptop - 6GB Memory - 750GB Hard Drive - Modern Silver

HP - ELITE 15.6" Laptop - 8GB Memory - 180GB Solid State Drive