Monday, July 30

Are Amazon's book recommendations getting spammy?

I'm a big fan of Amazon's book recommendations. For years, I've relied on them because they're the collective judgment of people like me -- people who have bought similar books.

So when Amazon asks me to consider a new book, I listen. In fact, I buy most of my new books this way. And believe me, I buy tons of books. Literally.

Sure, sometimes Amazon spits out a wacky recommendation. Like last year, when they asked me to buy a book about hysterectomies. Being a guy, I just wasn't interested. But hey, nothing's perfect, not even Amazon's computers.

Recently, though, I feel something has gone seriously wrong. What got my attention was this book recommendation I received in an Amazon e-mail last week:
As someone who has purchased or rated books by Robert A. Wilson, you might like to know that Mexico And Its Religion: With Incidents Of Travel In That Country During Parts Of Years 1861-64 And Historical Notices Of Events Connected With Places Visited is now available. You can order yours for just $34.95 by following the link below.

Mexico And Its Religion: With Incidents Of Travel In That Country During Parts Of Years 1861-64 And Historical Notices Of Events Connected With Places Visited
Huh? I've never bought or rated a book by Robert A. Wilson. In fact, I've never bought a book on Amazon by anyone named Wilson. And I've never bought a book about Mexico or religious history, either.

But here's what's really strange: Normally, Amazon recommends new books. But this isn't a new book at all. In fact, it seems this book isn't even being marketed by the publisher.

What am I talking about? Look at the book's product page on Amazon. There's no cover image, excerpts, or even a description of the book. No reviews, no page count, no blurbs, nothing! Why not? My hunch is the publisher figures it's so unlikely that anyone will buy this book, it's not worth the trouble. Even the publisher doesn't trust the book. But it's good enough for my Amazon recommendations?

What's going on? According to this account, (thanks, Denny) the publisher is systematically "swiping copyrights." He says it's reprinting old books without even asking permission from copyright owners. Sounds like a great business model. You get your content free by scanning an old book. You use print-on-demand publishing, so you have no printing costs. At least, no printing costs until some poor schmuck orders a copy.

But what really irks me is how Amazon is involved. How can I trust their recommendations anymore? Their whole value was unbiased recommendations based on other shoppers. Now it seems Amazon's recommendations aren't only advertising, they're random, irrelevant advertising. In a word, spam.

For all I know, this publisher is paying Amazon to send out zillions of these spam book recommendations, or perhaps Amazon has a stake in this outfit. I don't know. All I know is, something that was truly special about Amazon -- book recommendations based on what other people like me bought -- has been corrupted.
---------------
If you like my articles, click the orange button to subscribe:

Steve Weber is author of Plug Your Book! Online Book Marketing for Authors

Labels: , ,

4 Comments:

Anonymous Chad said...

This appears to be confuting several different issues. Kessinger's business is in reprinting (in ugly yellow covers) old books that are no longer in copyright. In most cases this is simple to determine; in others, it isn't, and no doubt some mistakes are made when you're talking hundreds if not thousands of titles. Robert A. Wilson's book was published in 1855; it's easily downloaded as a Gutenberg etext, so is unlikely to be a victim of "copyright swiping." Kessinger started their business long before the current POD trend, and I don't doubt that they might have lax standards & practices in copyright research, but anyone who thinks that a reprint of an old book is going to automatically draw sales is out of touch with reality. (And calling it "systematically" stealing material requires more than a single copyright owner's claims - have you documented such?)

Kessinger's model appears to be based on the long, long tail model. (Assuming set-up costs of $25 a book with LSI, they don't need many sales a year to cover expenses.) Personally, I think their reprints are poorly designed, and just make things more difficult for reprinters like myself -- particularly when seeking out titles that aren't yet reprinted -- but I seriously doubt they're in cahoots with Amazon.

More than likely, the book recommendation algorithm was being tweaked or had a glitch. (Or it may have something to do with Amazon confusing this author with Robert Anton Wilson, who appears in the Similar Items category.) In any case, the contention that this book is not "new," is incorrect. The published date is June 25, 2007. That qualifies it as a new reprint.

7/30/2007  
Anonymous Anonymous said...

I agree with Steve. It is another hurdle. Many of the out of print books I sell are for research. Those people want the information and will pay the lower price. The people buying on behalf of government learning institutions will be required to purchase the lower ones.

How exactly do the POD printers get past copyright laws? Are the copyright owners benefitting? It would be interesting and perhaps beneficial for us to know the way the system work, so that we would know which books are subject to reprints.

7/31/2007  
Anonymous Anonymous said...

I have had students mostly and some researchers email me with requests, not on the condition of the volume but on the information contained within specific chapters or indices. I usually reply with a PayPal invoice for a minimum of 2 hours at a fee from $15 to $75 per hour with no guarentee the information will answer their questions or of be of any value to their research. To date I have had no takers.

It does upset me that with regard to students, the main request seems to be Title Page information. Obviously, to pad a thesis with more references. We should all shudder at this practice if carried into their future professions.

7/31/2007  
Anonymous Newt Barrett said...

Steve,
You are right on the money in terms of spam-like recommendations.
Roughly 4-6 weeks ago, I started to get recommendations from Amazon that didn't match anything I had ever ordered--or would be inclined to order.
In fact, this morning I got one for an anti-Bush screed, that couldn't have been less applicable to me. Believe it or not, this applies to music as well in terms of junk stuff I have received.
Something is clearly amiss.

8/01/2007  

Post a Comment

Links to this post:

Create a Link

<< Home