If only Freud were still around—and interested in the publishing business. What do readers want? he'd ask. And then the corollary: Why do they go for some books more than others?

It is the universal question for book people, and the WSJ explored it in a front-page story last week about the diverse fates of two novels: Holt's The Interpretation of Murder by Jed Rubenfeld, and Atria's The Thirteenth Tale by Diane Setterfield. Both books are intelligent first novels with commercial appeal; each was highly sought, then highly touted by their respective publishers and well-received by booksellers. And yet, The Interpretation of Murder—for which Holt paid a reported $800,000 for North American rights—has, to put it mildly, not performed to expectation. The Thirteenth Tale, on the other hand, debuted at the top of several bestseller lists.

It was a fascinating story, extremely detailed and thorough in its analysis of how a book gets to market and "kills"—or not. Booksellers were quoted, explanations given—of Holt's choice, for example, to preempt with a North American rights—only offer (which, in hindsight, might have been a mistake, since the cost of an all-rights buy could have significantly defrayed even a much higher price because the book sold to publishers all over the world as well as to the movies); journalist Jeffrey Trachtenberg even addressed the oft-held notion that editors don't edit. (Holt's John Sterling clearly did.) It was even scary in what it revealed about publishers' motivations: Holt admitted that it thought Interpretation of Murder had Da Vinci Code potential. Haven't we learned yet that you can't count on miracles happening twice? (See Charles Frazier's $8-million Thirteen Moons, which is no Cold Mountain, or the fact that Little, Brown has been attempting to copycat its Lovely Bones success for years.)

Still, the story was fair and enlightening as far as it went—except that it didn't take into account one major factor in book publishing: word of mouth, or as former Warner head Larry Kirshbaum used to call it, "the fairy dust" that makes some books work. Sure, it's good to have a strong manuscript, and smart marketing certainly doesn't hurt. But what all publishers know, but hate to admit, is that you can do everything right—edit, market, promote—and still get relatively nowhere. For every The Historian, in other words, which Kirshbaum's marketing machine pushed to bestsellerdom perhaps greater than it deserved, there are the sleepers (Angela's Ashes, say, or The Kite Runner) that seem to come from nowhere. "What this shows," the writer Caleb Carr told the Journal, "is that you can't schedule cosmic events."

Put another way: what draws a reader to a work of fiction is intensely personal and not at all prescriptive; what makes that reader pass news of the book to others is as unknowable and indefinable as love.

Then again, all readers, like all lovers, are not created equal. Perhaps the most revealing comment in the Journal came from Barnes & Noble honcho Bob Wietrak, who said his executives "really liked" Murder but "fell in love" with The Thirteenth Tale and convinced 40,000 booksellers to talk it up.

If that's not word-of-mouth supersized, I don't know what is.

Agree? Disagree? Tell us at www.publishersweekly.com/saranelson