According to Amazon's recent annual 10-K filing, the company leased and owned (most of their space is leased) about 288,419,000 square feet of space around the world at the end of 2018. Of this number, about 80% is used for "fulfillment, data centers, and other." Amazon doesn't break out this line item any further, but GeekWire reckons that a good 3/4 of their real estate is dedicated to their fulfillment warehouses.

Here's the full summary of their facilities (from the 10-K filing):

Given that fulfillment is such a large share of their properties, I am most interested in understanding the geography of their warehouses and how that impacts their core value proposition, which is largely all about convenience.

In April 2017, Jean-François Houde (of Cornell), Peter Newberry (of Penn State), and Katja Seim (of UPenn) published a paper on this very topic called, "Economies of Density in E-Commerce: A Study of Amazon’s Fulfillment Center Network." There's also this Knowledge@Wharton podcast on the paper if you're looking for a quicker listen or read.

In the early days of online retail, the decision of where to warehouse had meaningful tax implications. Because (in most cases in the US?) you only had to collect sales tax if you had a physical presence in the same location as your purchasers.

As that changed, it then made more sense to create a broader distribution network and minimize the distance between fulfillment center and purchaser. By 2016, Bloomberg estimated that nearly 78 million Americans lived in a zip code where Amazon offered free same-dame delivery. That number has obviously increased since.

And in the paper "Economies of Density", they discovered the following cost savings as a result of Amazon's growing fulfillment network:

We find that Amazon saves between $0.17 and $0.47 for every 100-mile reduction in the distance of shipping goods worth $30. In the context of its distribution network expansion, this estimate implies that Amazon has reduced its total shipping cost by over 50% and increased its profit margin by between 5 and 14% since 2006. Separately, we demonstrate that prices on Amazon have fallen by approximately 40% over the same period, suggesting that a significant share of the cost savings have been passed on to consumers.

The interesting question for real estate people and city builders -- which is brought up in the Knowledge@Wharton podcast but is difficult to answer -- is whether there are diminishing returns to this "economies of density" phenomenon. In other words, how dense does Amazon's fulfillment network want to be?