Goodreads' data is…not good. So I've returned to one of the oldest tricks in the book: good old scraping - and actually end up with pretty decent data.