You write for magazines, blogs and newspapers — what do you like about books?

Carl Zimmer advises getting a good agent to help navigate the culture of publishing. Credit: B. STECHSCHULTE

They make a deep impact on people. A person is with you for hundreds of pages, and that's a long way to go together. There are cases where people became a scientist because they read a book of mine in college. That's intense.

How do you pick a topic?

I choose subjects that other people haven't written about. I don't want to do another iteration of a book that has been written many times over — I don't find that satisfying. It has to be something that I'm fascinated by, because this is going to be a year or two of my life, and if I start getting bored that's going to be a miserable experience. Sometimes I hear ideas for books that are based on obscure chapters in the history of science, but those stories are not interesting to anyone except in that field. Compelling people are also essential.

Is it a good idea for writers to chase trends?

Ultimately, I think it is a losing game. Publishing goes in waves. Freakonomics, for example, was a very successful book, and all of a sudden every publisher wanted their own Freakonomics-like book. Some did well, and some not so well, and now the publishers want something else.

How do you write a proposal for a popular-science book?

A good proposal reads like a magazine article. It is going to pull in a reader and get them excited about a subject. The proposal should give a good feel for the subject, explain why an editor wants to publish a book like this, why it is important, how it compares to other books and why you are the right author.

Who do you send it to?

Get an agent. The publishing world is a peculiar culture with customs that you can't figure out on your own. Look at books that have come out recently and look in the acknowledgement section. If an agent did his or her job right, then an author is going to thank them profusely for all the things that they did.

Will an agent take a chance on an unpublished author?

To become a writer, scientists do not have to have a long paper trail. It helps if they have been in the news themselves for their research. If you are a leading expert in a certain area, that can be compelling.

What should scientists who want to write a book know about the process?

It is a lot of work. They may have to write at night and drop out of teaching for a while. A lot of scientists think 'well, this is easy; this isn't scientific research, this is just writing for non-scientists'. But if you come back a week before the manuscript is due and ask them how it is going, 'easy' is not the word they're going to be using.

How do you manage the writing process?

Sometimes I'll write everything out on paper and post it on the wall as an outline. It is nice to see a book in the big picture because so often you're zeroed in on one paragraph. You have to be ready and willing to write a lot, every day. In the last few months before a book is due, I clear everything else out, and I may be writing for 10–12 hours a day.

What is the editing process like?

A good editor will actually edit a book. Unfortunately, in this publishing climate, a lot of editors function as acquiring editors, and don't put that much work into the editing process. Sometimes authors find that they have to hire freelance editors to get the guidance they need.

How much money will you make?

You may get a modest advance if they don't know who you are ... maybe US$10,000–50,000. If publishers think it is going to be a best-seller and they compete with each other to get your book, you'll get a huge advance. Anything above $100,000 means that people are really taking notice. But that doesn't last you as long as you might think. They pay it out in thirds, and sometimes quarters — part on signing, part on delivery, part on acceptance, sometimes part on publication — and your agent will get 15%.

What can a writer do to help sell their books?

You can't assume that you turn in your book and your work is done. You know more about how to market your book than your publicist. You know the best science journals to get it reviewed, the places you should speak, the blogs that people are reading.

What's your top tip for a researcher setting out on this path?

Find a subject you love so much that you could spend two years writing about it — even if nobody bought the book when you were done. That way, if people do buy the book it's a bonus. Love for your topic is very important.