The book Nightwork is about the history of hacking at MIT. Before the term hacking became associated with computers, MIT undergraduates used it to describe any activity that took their minds off studying, suggested an unusual solution to a technical problem, or generally fostered nondestructive mischief.
I have been thinking up an idea for a little nondestructive mischief. Amazon has a ranking for all their books. The first bestselling book, second bestselling book, etc... For example, the Nightwork book mentioned above is ranked 16,879. (At least is was when I wrote this) Books in the top 100,000 Amazon must sell a few copies everyday because a friend of mine that sells books through Amazon has noticed that books with a ranking of 100,000 or less sell relatively quickly. What I thought would be an interesting bit of hacking would be to select some book on Amazon that has a ranking of more than 100,000 and by sheer word of mouth, blogs, and email move the book down to the top 100 on Amazon. One thing that would make this more interesting is to select an interesting book to try this on. Anybody have any suggestions of a book that is ranked higher than 100,000 that has some unique qualities that would make it appropriate for an experiment like this?


I suspect you are going to need more information.

How frequently is the ranking updated? Over what length of time are sales averaged to produce the ranking? How long would a deliberate spike in ranking last? How many extra orders will it take to spike something up precipitously? Will that spike be noticed by anyone (who wasn't already warned to look for it)? I suspect that some of these questions would need an Amazon insider to answer them.

How could you organize such an effort (via the internet, presumably) and not have Amazon (a company that is VERY tuned to the internet) not notice? Do used books count toward the rankings? If so, perhaps you could organize a timed set of purchases of fictitious cheap used copies between several (many? a huge number of?) co-conspiritors. It would still cost a bit, as Amazon not only takes it's 15% of the purchase price on used sales, but also takes a piece of the postage/handling fee as well. Sure, 15% of $0.01 is $0 for Amazon, but the buyer still pays $3.49 for basic shipping of which the seller only receives $2.26 to cover postage. Even if you pull off a balanced A sells to B sells to C sells to D sells to A circuit, that will still cost $1.23 for every "purchase", for however many it would take for that title to experience the desired ranking spike.

Before all that, you should check with Amazon policies/terms. There might be a specific prohibition against attempts to game the ranking system.

How could you organize such an effort (via the internet, presumably) and not have Amazon (a company that is VERY tuned to the internet) not notice?

I was going to have people buy the books so I don't think it really knows if Amazon knows. The hack I was thinking of was not to trick Amazon put to take a book that was in the 100,000s and get soem buzz going about it to cause enough copies to sell to get it down to the top 100. You are corrrect that it would be useful to know how many books it would take to get into that range. That is one reason I was asking about books that people reccomended. If someone knew of a sleeper book that should really be a bestseller it might be easier to get a buzz around it and get people buying it that trying to push the book "Basics of Entymology" to number one. You did give me some additional ideas to think about.

Amazon sales rankings certainly measure the buzz on a book, but I'm not sure it has any kind of driving effect on it. I don't see it being used anywhere outside of the book detail page. Sure, Amazon uses the ranking to compile it's bestsellers list. But even in a niche sub-category, like Top Sellers > Books > Subjects > Medicine > Nursing, the #10 book has a sales rank of under 5000 and the #20 book has a rank of under 11000. There's gonna have to be a lot of sales to break far enough into any list that will be noticed widely enough to drive additional sales.

The bottom of the bestseller pages claim that the lists are updated hourly, which suggests near-instantaneous rankings, though I suspect they use a moving average over some more reasonable length of time, say a few days, to normalize out any noisy spikes.

I'd think you'd need to tap into something that has a stronger role in a popularity feedback loop. Kinda like the way bloggers have teamed up to spike Google with novel associations between search terms. (

Try my book, "Reliving the Civil War: A Reenactor's Handbook." When new, this was in the 127,000 range on, but has since dropped down to the 367,000 range. Still in print, it describes the method of turning an academic subject, history, into an active sport or hobby.
It is also available through as well, so this would be an interesting examination of the sales between a paper copy for sale, and an e-copy available from most large libraries for free.

Mr. Hadden's book can be found here.

Subscribe to Comments for "Nightwork"