Cameras tracking shoppers: the economics of retail video surveillance

We examine whether video surveillance data enables improved retail performance. We analyze sales data collected from six sporting goods stores. We relate the sales data to an experiment where three stores had placebo video surveillance, and the other three stores made use of information learned in the video surveillance starting half-way through the sample period. A difference-in-differences analysis of the data indicates the use of video surveillance enables a substantial increase in total sales per hour, the number of transactions per hour, and the average size of each transaction.