There are several things to consider before you sit down to write a book. But here are 10 I wanted to share because I think they are either very important or not spoken of enough.I've gone down this road many times, but I took a long break between books and instead focused on other areas. Of course I am talking about writing about technology, not Harry Potter #8.
I am currently in the early stages of writing my new book tentatively titled "Data Access with Silverlight 2". The publishing has not changed much other than that the number of books sold has decreased significantly over the years. In the late 1990's and in to 2000 many decent technical books were selling 15,000 to $25,000 copies. Today most never reach 10,000 copies sold. It was not a money maker back then and certainly is not now. Which leads me to rule #1
NOTE: Yes, these are completely arbitrary rules, but hey, its my blog so I make the rules today :)
Rule #1) Don't write the book with the goal being to make money. You may hit the jackpot, but very few books do.
Rule #2) Spend a good amount of brain cycles on writing the book proposal. Not only will your publisher appreciate it, but it will help you clarify the purpose of your book, the flow, and what should be included and excluded.
Rule #3) Keep the content focused. So many books seem to have everything under the sun between its covers. Let's face it, who wants a 1000 page brick on their desk (unless its a reference only)?
Rule #4) Use less words to say more. use diagrams, listings, code examples, and other resources to clarify your points.
Rule #5) Pick good reviewers. This is critical because as the author you will miss problems because you are so close to the material. A good reviewer needs to know the technical material, know how to communicate the problem to you, and most of all ... s/he needs to be available to dedicate some time to help review your book. So choose more than 1 :) And be prepared for making changes based on suggestions.
Rule #6) Interview the publisher. Talk to several publishers about your ideas. Find out what they think of it, how they would pitch it, review what other books similar to yours they have done. Make sure the publisher believes in your book.
Rule #7) Do something original. Create a book that grabs the prospective reader's interest and keeps it. The content is the ultimate means to persuade the reader, so keep it focused and create a 1 paragraph "why you should read this book" snippet. If you do your proposal well, you will have this at hand.
Rule #8) Expect long nights, a lot of effort, and frustration. You know your topic well, that's what made you an expert in the field. But do you know everything about a topic? Of course not. When writing a book you will learn things you did not know previously since you are now covering more of the topic than before. Drink it up, but be prepared for the possible delays that may occur because of it.
Rule #9) Write a chapter outline, keep notes, do research, organize your thoughts. I use OneNote 2007 to keep all of my notes on a book. I have notebook sections for each chapter and I jot down notes about what I want to cover in words. I also keep a bullet list of topics I want to hit heavily and a bullet list of examples i want to go through. Having a tool to do this is awesome. OneNote is ideal, for me. I used to use paper to do this :). But just organizing yourself really helps keep the flow moving.
Rule #10) Enjoy it. The book and the readers who enjoy it are the ultimate reward. Enjoy the process as much as you can. If it is fun to write it, it will often read better, too.