On the social networking site authonomy, I noticed something fast. Authors (including me) seldom know what's going to sell. The ratings for my books on that site were inversely proportional to the real sales. Nonetheless, over the course of your career as a writer, you can use feedback to adapt.
The 5 star method Amazon and Goodreads uses doesn't mean much to an author either. It doesn't tell us what we need to change to make a better product. Of course, the review contents can be helpful, but only if the criticism is specific and the audience wasn't misidentified. Indeed, what the current review system tells me most is, "Did I hit my audience with the pitch and categories?" Someone once mis-tagged one of my books BDSM, and those people are quick to complain when you don't show them what they're expecting! The same is true of any genre. A book I sent to a romance blog earned a D rating, but she admitted she couldn't put it down. When I moved the book to action + high-tech, it hit the top 40.
However, steering your writing career on this alone is like picking a church by the length of its name. Here are some other unusual metrics that tell me things.
1) How much does it sell in the first day of the giveaway? The second day, if you're in the top ten in your category, people will "me too". The first day tells you your curb appeal, or hit ratio: gawkers to buyers, the thing that will sell the most of your book. As a rule of thumb, the number of free copies (on a Monday for my genres) correlates to how many copies a book will sell in the first year, unless it hits one of the top hundred lists. My goal is to tweak the pitch (or even cover) until the sales exceed 600 and I hit the free top ten in my category before 5:00 pm. Otherwise, the book will never pay for itself.
2) How many people who read this book rate it? I'm not talking the professional reviewers (1 percent of those contacted) or Amazon giveaways (those people are like one-night stands who never call), I mean actual sales. My favorite book I wrote gets about one review for every 14 purchases. I won't list which of my books I pulled these stats from, but the review rate (in my opinion) tells you the true number of stars--how does it spur people to respond and share with others.
5 stars = 1 out of every 13
4 stars = 1 out of every 50
3 stars = 1 out of ever 200
Notice the linear pattern? Now, I want to say something important. There is nothing wrong with a professionally-polished three star book. With the proper marketing, it can feed you nicely, but it's cotton candy, not "To Kill a Mockingbird."
3) How many people buy the sequel? Because, abashedly, the only way I know of as a mid-level writer to make a living writing is with the series effect. In my genres, a series will sell 10 to 14 times more than a standalone. Plan accordingly. The quality of the experience with book one will correlate to how many people buy the next. Your mileage may vary.
5 stars = almost 100 percent comeback, and they read your other books too!
4 stars = 70 percent comeback.
3 stars = 25 percent comeback.
It doesn't matter much if number three in the series is a filet if number one is and overcooked hamburger.
4) Returned books tell you something when there's no free giveaway that month--probably how good your hook is or how well you represented the true story in your pitch. If you have more than one return a month for a title, look for something in the first three chapters that gave someone the excuse to say no.