I've got yet another reader question on writing (and I'm running out, so if there's something you want to know, ask). This question was about researching a book.
In my process, there are two kinds of research. There's the research I do before I start writing to help build the world and shape the plot and characters, and there's the research I do during writing or after the first draft to fill in the details.
Before I start writing a book, I like to do a lot of research on the setting and anything that might be appropriate for the situation. At this point, I have a vague idea of what the story's about and who the characters might be, and this research helps me firm it up. This is usually when I do any research that involves travel. Before I started writing my Enchanted, Inc. series, when I had just a general sense of what the story was about, I took a trip to New York and did a ton of walking just to soak up the atmosphere and to narrow in on where I was setting it. Buildings seemed to jump out at me as places where things might happen. Once I figured out approximately where the characters lived and where they worked, I took the subway routes and I walked it and timed it. As I walked around the city, the story and the characters really took shape in my head, so I was ready to start writing when I got home.
I also read a lot of books on topics that might come up in the story. Since I was basing the general idea of the magical corporation on the software industry, I read memoirs by people who worked in that industry and books on the various corporations. I don't use everything I read. I'd had this general idea of a difficult environment where a woman would struggle against the old-boy network and read some books about that topic, but it ended up not fitting when I started writing. I may also read things that might help me get into characters' heads. One of the characters in my upcoming new series is a police detective, and I read a lot of cop memoirs in preparation for this book. I wasn't looking for specific facts, more the general mindset and the way that kind of job might shape the way someone thinks, and that then influenced the character I developed.
Then once I have a more detailed plot and more specific characters, there may be things I need to research more directly that are good to know before I start writing. I've seen a lot of writers tripped up by legal system issues, for instance, or something like what's required to get a marriage license in that particular state. If you know your plot is going to have your characters get married, you need to know that. I've known of at least one romance novel that fell apart because much of the plot revolved around finding a sneaky way to get all the requirements for the marriage license in secret, only it turned out that most of those things weren't required in that jurisdiction. If the author had looked that up before she started writing, she wouldn't have had to rewrite most of the book.
Once you're writing, there may be things that pop up along the way that you need to know or figure out. You may need a specific idea for a restaurant for the characters to visit or a dish they might eat. You may need a train or airplane schedule to see if it's physically possible for them to get from point A to point B or to properly strand them. This is where you have to make a judgment call. Do you absolutely need to get this straight before you can continue writing, or can you fill in the blanks later, once you've finished a draft?
That's because research makes an excellent procrastination tool. If you aren't sure what happens next and you're feeling stuck or if you're just tired of writing and want to play on the Internet, then research allows you to take a break and still feel like you're working. Never mind that you're getting increasingly sidetracked down the Wikipedia rabbit trail. But research can also help with the good kind of procrastination, allowing you to work things out in your head before you plunge into writing. I've found that if I just jump into writing the moment I come up with an idea, it tends to fizzle out. The more time I spend in pre-book research, simultaneously absorbing information relevant to the story and letting my subconscious play with the idea, the more successful the book seems to be. The trick is in knowing when you're just stalling and when the research is actually necessary, and that's a fine and very fuzzy line. For during-the-book research, the deciding factor is whether that detail will affect the plot. If it's just about the color of the bellman's uniform at a particular hotel, I can flag it and look it up between drafts. If it's whether or not it's possible to get a train from point A to point B at that particular time, then that will affect the way things happen, so I need to know before I write that part and then either have to change plans or change the set-up.
Research really is important because it's often factual errors that make readers give up on a book. It's hard to make readers believe the stuff you make up if you don't get the "real" stuff right or if you don't manage to make things feel real. We all know it's fiction, but while we're reading a book, we like to have the illusion that it could really happen.