In her post below Lyda poses the question:
But it does seem to work for some people on some level, and I always end up wondering by what magic is that done?
I think it's pretty straightforward actually, and it all comes down to that word seem. Here's how I think it works (all numbers made up).
If fifty percent of all authors do self-promotion, and a random six percent of all authors cross over into best-sellerdom than three percent of authors who do lots of self-promotion are going to cross over into best-sellerdom purely by chance. Then, at least some of those authors are going to figure that it was self-promotion which made the difference whether it had anything to do with it or not.
Likewise, if you're watching from the outside, you might think the only thing that differentiates them from the herd is that self-promotion, and then leap to the same conclusion. For that matter, I will even concede that some particularly clever bit of self-promotion that hasn't already been done a bunch of times might catch the mood and go viral, but I think that's as much a form of luck as having the book do the same thing.
Great books with tons of self-promotion die. Barely adequate books that get very little push become best sellers. Most of the difference there is luck in hitting the right literary kink for the moment.
We want the industry to make sense, so we tell ourselves stories--we're authors, telling stories is what we do. That book did so well because the author came up with the really awesome book trailer. That one did poorly because the cover sucked. This one over here is a best-seller simply because it's that good.
But the truth is, nobody knows what's going to make a book take off. If there was a real answer, there's be a publisher somewhere that didn't sell anything but best sellers.