How a publisher can MAKE a bestseller - if they want
In hardcover, Elizabeth Gilbert's EAT, PRAY, LOVE sold 25,000 copies. The Wall Street Journal calls that mid-list. I'd call it pretty damn good. Anything over 10,000 is pretty good. The WSJ also looks at how "the book's transformation from respectable-selling hardcover to paperback sensation was no accident. It came about after a series of calculated moves from Viking's sister Penguin paperback line, where executives worked to interpret sales patterns and create a marketing blitz to attract individual readers as well as book clubs. Penguin's approach shows how publishers, which typically don't conduct market research, are becoming increasingly adept at hand-picking certain titles for stardom."
[Full disclosure: pick me! Pick me!]
And here's another sad truth from the article: "The vast majority of books face a tough reality. New releases that fail to take off in the first couple of weeks -- when publishers often pay to place copies on stores' front tables -- are relegated to the back shelves. 'The usual reflex is to give a book two weeks in the sun and then move on,' says Bob Miller, president of Walt Disney Co.'s Hyperion book division."
[Full disclosure number two: YA seems to have a longer window, which lasts until it wins some awards - or doesn't.]