Here is something I've seen people asking about: Print runs and what's considered a success. Like, how do you determine a print run and what makes a book a success.
Well, here are the numbers:
Somewhere in the neighborhood of 300,000-400,000 books come out each year in the US. And 95% of books that are released will sell fewer than 5000 copies.
Here's the deal: publishers estimate the number of sales (more on how to do this later) that they will make from a title and try to get close when they do a print run. If they have too many, the books will cost them money to store and will eventually be pulped. If they underestimate the demand, they will have to do a second print run, which costs more than having all the copies printed when the machines are set up the first time.
However, if there's a higher demand for a book, the publisher is usually happy to do another print run. It's also WONDERFUL for the author's professional reputation. If an author's first book has an initial print run of 5000 and sells out, that's better than having a 10,000 unit print run and only selling 6000 copies.
So, selling out your initial print run is one measure of success.
Another measure is getting onto a bestseller list. There are a bunch, but the most famous is probably the New York Times's. These are usually compiled weekly (Amazon updates every hour or so), and getting on one usually feeds into increased visibility and therefore increased sales, so it can be a feedback loop of awesomeness.
So, how does a publisher decide how many copies to print? Well, we rely on several things, such as information on perceived demand from our distributors and from advance orders. We can also make rough estimates based on online buzz on sites like Goodreads and pre-orders on Amazon, knowing what the numbers were like for similar titles in previous years and how well they sold when they were released.