The amazon Myth

[Karen Dionne]

I'm reposting my comment from Sandra's discussion below, because while the idea that amazon is NOT an important sales outlet for traditionally published books is common knowledge in my particular writers' circle, some of the newer authors here might not realize that amazon sales (and by extension the site's rankings) are insignificant.

A couple of years ago, a New York Times best-selling author -- someone at the top of their game, whose novels consistently debut on the list -- shared their actual sales figures with Sarah Weiman. Check out this excerpt from Sarah's Confessions of an Idiosyncratic Mind:

Over a three week period this summer, the following sales numbers were recorded for a NYT bestselling thriller writer's most recent book:

B&N: 4,140
Waldenbooks: 4,888
Borders: 3,993
Anderson Merchandisers/Walmart: 47,671
Target: 16,341
Price/Costco: 17,291
Sam's: 14,108
Amazon: 320

Sarah adds:

I'm not sure what shocked me more: the unbelievably low number for Amazon, or just how powerful Walmart and Costco really are in the publishing business.

The author further adds:

For all their hype, the truth (and I've seen this with actual sales figures going back to 2000) is that Amazon numbers are tiny compared to virtually every other retail outlet.

Amazon makes their profit selling used books, not new ones. Maybe their low sales numbers was one of the determining factors to shift their focus toward used sales -- I don't know. But I do know that their numbers are insignificant to the pub in determining the success/failure of a book.

Surprised that some of the figures are so "low?" Bear in mind that a huge percentage of actual retail sales are from independents, grocery stores, pharmacies, outlets like that -- which don't report weekly numbers.

But the Walmart number is rather staggering, isn't it? It's one reason I put what little local/regional promotional efforts I do into cultivating Anderson reps and going on day-long road trips to sign and sticker stock for Walmarts. They get a hell of a lot of foot traffic, and sell a hell of a lot of books. [end quote]

My take goes like this:

Since amazon barely sells books, all their rankings really amount to is a marketing ploy directed at writers. If they can get thousands of authors visiting the site compulsively to check their amazon ranking, it boosts amazon's hit count and brings in advertising and makes them look more important in the publishing industry than they really are.

When my book comes out next year, I've vowed NOT to look at my amazon numbers -- not ever never. Why should I care how I'm ranked internally based on sales figures by a company that sells an insignificant number of books?

Now if Wal-Mart had a ranking system . . . . . .

.

[Newt Love]

Statistics are interesting, but only relevant if the sample is adequately qualified and matched to other populations of interest. (Yes, I have a math degree, and know probability and statistics.)

All we know is that the book sales data presented above are for "a NYTimes best selling author." What are the demographics of the readers of that author? Could it be that the readers of Mary Higgins Clark are -- as a class -- less Internet savvy than the readers of Andrew Vachss? How tech savvy are the readers of science fiction? How likely are the loyal readers of literary fiction to run to Amazon.com to buy a book? How likely are folks over 40 to jump on the net to buy a book. Does the behavior of the over 40 group predict the Internet behavior of folks under 30? Who's buying your books?

If these sales figures are for a book that appeals to the types of readers that think the Internet is geeky and have never heard of Star Trek, then OF COURSE the Amazon.com sales figures will be puny. On the other hand, the sales of current NYTimes best seller -- Walter Isaacson's Einstein: His Life and Universe -- may have quite a few Amazon.com sales.

The first rule of statistical inference is that you can use the results of one study to infer something about other populations only as long as the two populations are similar. Without knowing the demographics of the readers in the first sample, we can not infer anything about any other population.

Please continue the discussion, but don't jump to conclusions based on a faulty understanding of how probability and statistics actually work.

[Bill Cameron]

Yes, my thought exactly. Furthermore, we're talking about a sample of one. A real study that broke down sales for a variety of different authors in many genres and sub-genres would be necessary before we could really develop any meaningful inferences.

[Karen Dionne]

Well, it's not just any NYTimes best-seller, it's a thriller author, which is one reason I thought the stats might be of interest to writers on this board (says so in the beginning of Sarah's post), but I hear what you're saying re this being just one example. Obviously, you can't extrapolate this data and say it's true of all books, and a statistical sampling of one is just that --

Still, I find these stats eye-opening. But then maybe that's because I'm a thriller author with a book coming out in mass market paperback next year from Berkley . . .

[Eric Christopherson]

Using research methods speak, I think what Karen is doing is presenting a "case study. "

Wouldn't be prudent to generalize the findings based on the data, but on the other hand, these data provide compelling reason to conduct a more ambitious research project using probability sampling so that inferences can appropriately be made.

Those are big differences in numbers in her little case study and given that a majority of Americans are now online, I doubt that one author's readership is going to differ much from another's on the internet savvy scale...

[Laura Benedict]

Statistically valid or no, this is still very interesting to me as someone just entering the sales market. I spend most of my day at the computer, and it is very easy to check up on Amazon to see what my book and others' books are doing. Amazon has, indeed, grown to mythic proportions in my mind. These quotes, Karen, and Newt's input give me a great measure of relief--if only to tell me that I don't actually have a good understanding of what I'm looking at!

[Jon Jordan]

I've actually heard from a number of different sources that the very highest percentage of sales on Amazon are 5%, and that's a high number, usually it's lower.

Worrying about sales on Amazon is really a waste. You would be much better off making friends with the independant books stores.

Which is why the smart authors don't link to Amazon to sell books.

[Laura Benedict]

On my novel's myspace page, I have a link to Booksense, which directs folks to local stores, but also to Amazon. Maybe I should drop the Amazon link and pick up B&N or BAMM since some folks find the Internet convenient.

One thing I do like about Amazon is the information that's available when I'm reviewing a writer that I haven't reviewed before--They almost always have good descriptions and Booklist or PW reviews of their other books, pub dates, etc.

Thanks for the advice, Jon!

[Barbara Fister]

There is also the Independent Mystery Booksellers Association which has a "find a store near you" feature. There are some great stores that aren't members, but still it's a place to start.

Powell's is another indie that is huge and has a good web presence. Their interviews rock.

[BrianLindenmuth]

Those figures are basically a snapshot that don't necessarily tell the entire picture.

True the powerhouse nature of a Wal-Mart cannot ever be denied but I think that Amazon's biggest strength lies in their long game. I say that based on one word and one word only.

Returnability.

For those that aren't in the book industry its simply the ability of a book seller or distributor to return a product at any time for any reason.

Now, all of your major players in the book selling industry use Returnability to their advantage to offset monies owed. Here's an example: Let's say that there are 1 million dollars of invoices that are currently owed, Company X will look at their inventories and sales trends to see what they have a surplus of and return a large quantity of product for credit. This credit will offset what is owed. Thus they will then pay less for the current invoices.

These companies do this with the full knowledge that there are no limitations to re-ordering. Meaning that they can return 17,000 copies of book Y one month then turn around and re-order the same title the very next month. It's basically a revolving door of product and money.

Now, this brief summary leads me to this:

Amazon has full returnability but hardly ever uses it. If Amazon orders a certain amount of copies of a book they will keep it till it gathers dust in the warehouse.

Amazon can be an authors best friend much later down the road, especially an author whose titles don't sell a large amount of copies. Long after other stores are telling a book buyer that a particular title is out of print, or a company went bankrupt or it will be a special order or whatever the case may be Amazon will have decent supply amounts for that title.

[norby]

I guess I'm not entirely certain what you're saying here. Is it that you shouldn't focus on the fact that Amazon sales numbers are so low? Then I agree-why get yourself upset?

But there seems to be this undertone that suggests if Amazon sales are so low, why bother with it? Maybe I'm reading it wrong-I hope so, because personally I buy about 95% of my books from Amazon. The closest bookstore in my town is on the other side of town, and it's filled with people who know nothing about mystery novels. Not to mention a manager who needs to be taught some basic business skills.

It's one of the big box stores, and the only other bookstore in town is one of their subsidiaries-not much choice really. So Amazon is my go-to. For books, movies, cameras...

[Karen Dionne]

Hi, Norby -

There's nothing wrong with buying books from amazon - I do too! What I'm talking about here is what happens when new writers obsess over their ranking on amazon, thinking (erroneously, I believe), that the ranking is a reflection of how well their book is doing as a whole. Since amazon sales make up just a small fraction of a particular book's overall sales, it's my opinion that authors shouldn't concern themselves at all with their amazon rank. A lot has been written about how a single sale can drastically raise the rank - which to me, lends support to the idea that amazon.com isn't a significant factor in overall sales.

Here's an excerpt from a Wall Street Journal article:

"Amazon says little about how it calculates its rankings, though scholars and publishers have attempted to reverse-engineer the system to determine how a sales ranking translates into actual sales. One major quirk: Used and new book sales are counted equally. So an author anxious about his sales ranking could put a few dozen of his books for sale for a penny apiece and ask a friend to buy them all.

"This all adds up to numbers that are ubiquitous, closely watched -- and of dubious value. The targeted marketing campaigns contribute volatility to sales-ranking numbers that are inherently unstable. Outside the top 1% or so of books, few sell multiple copies a day, so little separates books with rankings tens of thousands, or even hundreds of thousands, apart. Morris Rosenthal, an author and publisher based in Springfield, Mass., who has studied the Amazon charts, says a day without a sale can send a book ranked 10,000 to as low as 50,000."

Do readers even pay attention to a book's amazon ranking? I suspect not. And since it's not a true indicator of how well a book is doing in the market, I don't think authors should either, that's all.

[norby]

That's what I was wondering. I can understand the temptation to obsess about what the sales are at each outlet, but if you know that a certain retail option is usually lower for most authors, then it isn't worth stressing yourself more than you already are.

I know I don't look at book rankings-I go looking for the books I want or an author I'm looking for or, like for Mother's Day, something I think someone will like. I couldn't give a flip what they're ranked.

It's odd that used book sales would be counted with new book sales considering that used books are typically sold by secondary sellers through the marketplace (something I rarely, if ever, use). I guess a sold book is a sold book.