Several of our technical titles have sold well over 20,000 copies but we've also had the rare title sell in the 2-3,000 copy range. That is definitely rare and only on very niche topics.

Writing a technical book is a lot of work on both ends, especially when a publisher provides real editorial and marketing support. Unfortunately, most no longer do, which is unfortunate for all involved and a good argument for considering self-publishing. After all, if publishers aren't offering added services, why not self-publish?

We read and edit every word before a book goes to copyedit. I've personally spent well over 300 hours editing individual titles. If you were to hire someone to do that work the minimum cost would be $15,000.

I'd love to see more people writing technical books -- good technical books -- for love of the subject and out of a desire to share knowledge. Sure, a good book will build your brand, but make it great, first.

The world always needs more good books. Our watchwords are "fewer, better books." Aim to be great.

I guess you're selling so many copies of each book (and based on your watchwords) by being quite selective about what titles you publish?

For the record, I do think my book is great but writing a book is not black and white - it's such a big endeavor that there is generally more than one motivation. I also didn't want to self call.

The author was making the point that the good-case scenario for book sales results in a not-very-impressive royalty. And that implies the average-case will result in a really-not-impressive royalty as you've personally seen.

So we do need to buy more tech books if we would like to have authors compensated fairly... which will hopefully attract more talent and effort towards creating more good tech books.

$45K is good money as far as I'm concerned.