This is a blog post that I've actually been chipping away at for a couple of weeks. Invariably, writers need to decide what to name their creations. Sometimes this comes wholly naturally, while other times it is a struggle that goes on for days, weeks, months, until eventually a name is pasted on and it will just have to do. Naming is such an odd prospect; it needs to be one part art and one part science. It can be the method through which a great novel by a relative unknown can use a good title to separate itself from the chaff that comes out all too often. Similarly, a bad title can preclude someone from buying a book. If you're pitching your book the traditional route to publishing houses; that name may make the decision for the person who reads over your pitch.
Naming, however, has become less of an artistic whim and more based on numbers and purchasing patterns. Lulu.com recently ran a study that suggested that Sleeping Murder, a short, figurative title out performs a literal one like The Da Vinci Code. Their system has a 70% success rate at predicting what will be a best seller using a piece of software they've created, that you can reach here, which out performs most people. Conventional wisdom supports the idea that short titles are better, but studies haven't fully supported this. For nonfiction books, it seems to be more important to have a short title than it is for fiction.
One of the key things to do is to research your genre as much as you can. You need to know your genre before you can stand out in it. Take a look at the length, the words being used, and what they're like. Try to do you research to see which books are performing better or worse on different lists. It is always going to be a difficult issue to separate how much of the book's sales are good marketing, a good book, or a good title. True best sellers are going to be a mix of all of these things. You need to see if there are trends; are all the books literal or figurative now? You also need to see if those at the top of the list are different from the great bulk of books currently coming out. Tying the title to the cover, which should also be part of this research, can be huge. If your book is the only book that is a white background with a big, short title, in a sea of darker covers with longer titles, like Scalzi's recent book Lock-In, you're that much more likely to stick out in the crowd.
Finally; you need to do your research about the names you're deciding on. How likely are they to be associated with different concepts, ideas, or trends. What happens when you Google them? If you have someone who has been told by a friend that they should check out your book, and they do a search for it to see what reviews are like, will they see your book or reviews on a heavy metal band with the same name? These are all circumstances that can easily hurt your chances at sticking out in the crowd.
So good luck, do your research, and knock them dead.