I've heard the same about marketing and that was my impression when I was interacting with O'Reilly and Apress several years ago. For technical books today, they expect the author to bring the audience.
> I also don't get an advance.
Sure, but that's just a small loan. If you're in the economic bracket of a typical HN reader, you don't need this.
So what's the point to go with them? We all can save it as a pdf and upload to Gumroad,or even send it to Amazon to print. I appreciate books on JAVA or Docker don't sell in millions but I'd at least expect the publisher to try to put my book on the shelves..
Good question. :)
It depends a lot on what the author wants to get out of it and what they want to do themselves. Going with a traditional publisher gives you:
* The prestige of your book coming out of a well-known publisher. If your goal is mostly to boost your resume or public speaking fees, then having an O'Reilly animal on the cover of your book can be very valuable.
* They provide developmental editors, line editors, copy editors, illustrators, and typesetters for you. If you don't want to do that yourself or find freelancers, that saves you from the headache.
* Likewise, they deal with the administrivia and legal stuff around distribution, translations, ISBN numbers, licensing, etc. I imagine they might provide some legal cover for you if you, say, defame someone in your text or use a copyright image.
* Your editor will create a timetable and some urgency for you to get the book done. That can be really helpful for people who have trouble summoning the discipline and structure to complete a book.
All of that stuff is valuable. The question is whether it's valuable enough for them to take the lion's share of the profit that the book generates.
There are other people who focus on "low content" books like journals, log books, and children's activity books. They tend to more often do seasonal books.
I have done one "topical" book, about a new term that is trending on google over the last year, but it basically gets no sales. My mistake for not following the formula.
When a customer orders a book, you make 60% of the profit, minus the cost to print the book. The cost to print the book varies. My books run 30-40k words and cost between $3-4 to print. I sell the books between $11 and $14, so I make a few bucks from each book. It isn't a lot if you consider the price of the book, but it's better than traditional publishers by a LOT, and I don't have to print any books ahead of time. My upfront costs are just getting the books written and the covers made. Then I pay for ads on Amazon as a continuing cost. And I get complete insight into how well my books are selling on a day-by-day basis.
I miss out on things a traditional publisher can do, like marketing and real in depth book design, but from what I understand they don't do much marketing unless you're someone famous anyway. I also don't get an advance.
I do keyword research before creating a book, and I've looked at technical books, and he's right, they just don't sell very well. Maybe they used to, before stack overflow, but they're just not worth the investment. If you're thinking of writing one, do it only as a resume builder, or to build a community around something you love.