Michael Eisen, over on “It is NOT Junk,” has uncovered a pretty remarkable phenomenon on Amazon. Looking for an out-of-print book on developmental biology, he saw that Amazon had 17 copies for sale: 15 used from $35.54, and 2 new from $1,730,045.91 (+$3.99 shipping). Hmm. The next day, the price had gone UP — $2.8 million a copy (+$3.99 shipping).
One bizarrely-overpriced copy might be explicable as just an error – but two of them? From different sellers? What’s going on?
What appears to be going on is that the two sellers are using a form of algorithmic pricing – using automated systems to set book pricing based on others’ pricing. These two sellers happen to use each other’s price as the guide. Seller 1’s pricing rule for this particular book is: Set the price at 99.8% of Seller 2’s price. Seller 2’s pricing rule, though, is: Set the price at 127.059% of Seller 1’s price.
Seller 1’s strategy is easy to understand – but why would Seller 2 want its price to always be around 30% higher than Seller 1’s price? Eisen’s got a theory, and I think he’s right: Seller 2 doesn’t actually own the book – if you order it from them, they’re just going to buy it from Seller 1! So they need a cushion (27.059%) to compensate and to make sure that there’s still profit from the sale.
I’ve noticed before that Amazon itself appears to use some weird forms of algorithmic pricing – for instance, when I’d check the Amazon page for my book, I would notice odd fluctuations day-to-day, anywhere from $16.25 to $24.95; as best I could make out (not being as systematic about these things as Eisen is), the changes appear to be inversely related to the number of sales – as sales went up, the price would go down, while if nobody buys the book for a few days, the price drifts up. Seems counter-intuitive, though I suspect the Amazon folks know what they’re doing . . .