Jump to content

how do I begin to write a book

excitementinlife's Photo
Posted Dec 30 2009 08:47 AM

how do I start writing a book

2 Subscribe

7 Replies

+ 5
  blackbear's Photo
Posted Dec 31 2009 03:54 PM

That's a fairly open question, so I'll make a few assumptions here:

1) You're talking about a technical book, specifically a computer book, not the next Gone With the Wind.

2) You actually have an idea for a topic, rather than just generally wanting to write about something.

3) You have a relatively good level of proficiency in the topic, decent writing skills, etc.

Unlike fiction, where you typically have to have at least a chapter written before you shop the book around to publishers, non-fiction technical books are all about a table of contents and an outline for each chapter. So a good start is to lay out an outline at a moderate but not obsessive level of detail. So, for example, if my topic was taco making, I might start with:

1. The history of tacos. The first known taco. Tacos in pre-Incan America. The taco comes to Mexico. The introduction of tacos to the United States. Tacos today.

2. The basic ingredients: What makes a good taco. What types of cheese are used. Bean, chicken, or beef? Blue corn tacos, are they authentic.


(BTW, no assumption should be made as to my level of expertise in taco making, the previous examples was for illustrative purposes online. Professional Driver on Closed Course, Do Not Attempt.)

Anyway, you're also probably going to need to be able to answer the following questions:

1) Why are you the right person to write this book.

2) What is the potential market size for this book.

3) Are there competitive titles already out or in production? How are they selling.

4) What do you see as the rough size of the book, and how fast can you write it.

Every publisher is a bit different in their proposal submission process, at O'Reilly you'd start by going here. If there is interest, they'll get back to you and start refining the proposal. Just because interest is expressed doesn't mean you're home-free. There's usually a committee that meets to review proposals and green-light ones to move on to contracts. This can take a while (weeks to months).

Also be prepared to tune your proposal to different lines inside the publisher. For example, my first book (MySQL and JSP Web Applications, published by SAMS), started as a "Learn X in 7 Days" proposal, then was written as a "Learn X in 30 Days" outline, then finally tuned again to a general title book.

That's essentially the process through to actually starting to write. The actual process of writing it is an entirely different topic, and outside the scope of this answer.
James Turner

Contributing Editor, O'Reilly Media
Correspondent for the Christian Science Monitor
+ 1
  arjonesiii's Photo
Posted Jan 04 2010 07:21 AM

Blackbear gave a good answer. With all the same assumptions in mind, I'll just add a few things that you might consider.

1. Do you have a particular audience in mind? Are you writing a "beginner" book or are you targeting a more advanced audience? Audience matters--a lot. For example, you don't want to go through the history of your particular technology if you're writing for an advanced audience. After you decide on an audience, focus on what that particular audience needs, and write only that.

2. For each major section or chapter of your prospective book, can you come up with a realistic (or at least interesting) example? Readers are tired of books and articles full of "HelloWorld" examples. Sure, they show you a tiny bit of syntax, but they're also essentially useless in the real world.

3. Can you arrange the material you're trying to teach in a way that makes it easier to learn? In other words, how will you approach the topic pedagogically? Bear in mind that your goal is to get readers to your level of expertise more efficiently than you were able to get there, and preferably in a more interesting way. Can you identify likely errors and show people how to avoid them?

4. How much time do you have to write? Realistically, it takes several hours per day for about six months to write a decent tech book. If you don't have that kind of time, or don't want to make that kind of committment, consider writing articles instead.

With all that said, writing a book can be a great experience, and you'll likely find that *you* will learn far more than any prospective reader.

Good luck.
+ 1
  roger69's Photo
Posted Jan 04 2010 11:45 AM

The previous two answers contain lots of useful tips. I'm going to add one caution:

Don't expect to make any money.

If this is a technical book, you'll get an advance on royalties (generally about 8-10%). You'll need to sell enough to cover that advance before you make any more money. If this is a typical techinical book it's going to be about some niche that might be interesting to a particular audience for a period of time.

There are exceptions, of course. The O'Reilly Perl books are huge sellers and have been for years. In my own experience, the advance on my two books was money that I really needed at the time, but I was under no expectations on making any more.
Roger Weeks
Network deity or designated scapegoat,
whichever is more appropriate
+ 2
  MaryO's Photo
Posted Jan 08 2010 12:59 PM

You've already received great feedback, so I will keep this brief:

Check out http://oreilly.com/o...thor/index.html

Refine your elevator pitch based on the goals, approach, and promise of your planned book.
Read some books or websites about the process. Forest for the Trees has a few good chapters.

If it's a technical book, send me a synopsis and we can talk.
+ 2
  adfm's Photo
Posted Jan 11 2010 05:30 PM

I'm sure everyone approaches writing a book differently. If you've got something to say that you feel would make a good book, just start writing. You'll eventually meet your editor and they will help you get things straightened out.

Here are a few links to get you pointed in the right direction:

How to write a book – the short honest truth
by Scott Berkun

DIY: How to write a book
by Danny Choo

So You Want to Write a Book?
by Dave Thomas
  morville's Photo
Posted Jan 25 2010 09:04 AM

I start by sketching an outline, brainstorming titles, and thinking seriously about whether I really want to commit a year or more of my life to this topic. I may write an article or give a presentation to further develop my ideas and feelings about the subject, and to see what sorts of reactions and suggestions it evokes outside my own head.

Then, I spend about a year "thinking" about the book while mostly continuing with business as usual.

Sometimes, I decide not to write that book, and the cycle begins again.

If the idea survives, I eventually realize that my "thinking" has evolved into procrastinating.

That's when I contact Simon St. Laurent at O'Reilly, so he can make me actually write the damn thing!
Peter Morville
President, Semantic Studios
 : Mar 15 2010 05:34 PM
Out of curiosity, on the technical book topic, it's seems like it would be difficult to attract large audiences to material that can be read online, in a manual or that is part of a technology that changes rapidly. That hasn't stopped many a publisher from releasing title after title (there are some foundation concepts that change little), but in the case of technology like Java or Linux where most of what you need to learn the language or system is available online, or where there are often major shifts in the core of the technology, what is O'Reilly's approach to book proposals that are geared toward such topics? I would assume that the books would need to be tailored such that the content avoids JDK or Linux versions as much as possible, or libraries that are unstable in the examples. I'm also assuming the proposals would need to take a unique approach, not simply repeating what the man pages or info contain, or the latest Java Tutorials - no plagiarizing! I know I've been in the business a long time (about fifteen years) but I have this mental image each year of stacks of previous-version technical book on anything from Microsoft apps to Red Hat Linux tomes.
Christian Bryant,
UCLA JUG Founder
IT Project Manager