Selling Books with Algorithms

Abstract

In 1997 Amazon started as a small online bookseller. It is now the largest bookseller in the US and one of the largest companies in the world, due, in part, to its implementation of algorithms and access to user data. This Element explains how these algorithms work, and specifically how they recommend books and make them visible to readers. It argues that framing algorithms as felicitous or infelicitous allows us to reconsider the imagined authority of an algorithm's recommendation as a culturally situated performance. It also explores the material effects of bookselling algorithms on the forms of labor of the bookstore. The Element ends by considering future directions for research, arguing that the bookselling industry would benefit from an investment in algorithmic literacy.