Some of you may be unfamiliar with Amazon.com rankings. Amazon and Barnes & Noble both rank books according to relative sales position. The book ranked #1 is the vendor's top seller overall at that moment in time.
Amazon lists nearly 8 million books; if a book hasn't sold any copies since they instituted the ranking system, it just doesn't have a rank at all, so if you see a book with a rank in the four or five million range, you'll know it's a poor seller but not an utter washout (non-selling self-published titles apparently get booted out of Amazon's catalog after a time unless they were done through BookSurge). A book's rank is somewhat persistent; a book may gain a good rank due to a burst of sales, but if it doesn't sell any more after that, it will gradually fall down into the one to two million range. (The most poorly-ranked book I was able to find after a couple of Amazon searches is a 1988 academic education research book that has a current rank of #7,401,418.)
Both Amazon and B&N are pretty cagey about how they calculate their rankings; Amazon's rankings seem fairly volatile, with a single sale causing more movement on that site than on B&N's.
In general, a ranking in the five digits -- say 50,000 -- means your book is selling decently well on Amazon (it would also mean that 49,999 books are selling better that hour, but that over seven million are selling worse). A four-digit rank is really good, and if you break the top hundred for any length of time, you and your publisher should be jumping for joy.
What does an Amazon.com ranking mean in terms of actual sales on the site? Obviously, it's all fairly relative, but if your book jumps from 300,000 to 50,000 then that's a pretty clear sign that you made at least one sale, possibly more. A jump from 50,000 to 1,000 would mean a significant number of sales during the course of the day. Gaining a higher and higher ranking requires more and more sales; if you were to plot it, the numbers would probably make a J-shaped sales curve.
How does an Amazon.com ranking relate to sales from other vendors? Again, it depends. Some industry watchers I know say that for an average author who has a novel from a mass market publisher -- that is, a publisher whose releases are stocked by major brick-and-mortar bookstores if not Wal-Marts and grocery stores -- Amazon.com sales represent 3%-5% of overall sales. If you're a small-press author whose books don't get into most stores, or if you're an especially web-savvy mass market author who promotes heavily to readers who prefer to buy online, then your Amazon.com sales may represent 60% of the overall sales or even more. (Read more about different vendor statistics over at Morris Rosenthal North American Book Market.)
So the upshot is that being able to track your Amazon rank is a handy thing if you want to know how your book's sales are doing in very general terms.
My favorite site for tracking is Titlez.com ... it's free (so far) and you can track dozens of books. It appears that the site earns its keep for its maintainers through Amazon referral fees, but they reserve the right to start charging a fee in the future.
There are other sites that perform similar types of ranking and rank tracking: Sales Rank Express, RankTracer, Books & Writers, and Rankforest.
However, I haven't waded into them to see which one seems to work best, largely because only Sales Rank Express is completely free, and it provides only a book's current ranking with no tracking ... honestly, it doesn't really give you any more information than you'd get by simply going to the book's page on Amazon. Rankforest does allow you to track a single book for free, but past that it starts to get expensive. I'd be paying $19 a month to track 20 books, although I'd be able to track the rankings on Barnes & Noble as well as Amazon. Tracking 100 books would set me back $59 a month.
Given the listed monthly costs of the pay-for-tracking sites, it would seem the better bet to get a subscription at Publishers Marketplace, which also offers tracking for Amazon and Barnes & Noble (my agent tells me PM is pretty good for agents and editors, but he isn't sure if the service will be useful for most writers).
So, the upshot is that I keep hoping Titlez will remain free and functional. I don't think I'm really prepared to shell out $20+ a month for tracking statistics I can probably do on my own with a notebook and a calculator. And the curmudgienne in me tells me that if Titlez stays gone, that's one less site I'll feel compelled to visit during the course of my workday. And when you come down to it, we can all probably stand to do less web surfing and more writing.