Astronomy grad student James Davenport loves maps, data, and coffee.
He brought them all together last fall in a project he calls “The United States of Starbucks.”
Analyzing the locations of Starbucks in the lower 48 states, Davenport, who studies populations of stars at the University of Washington, turned up some interesting facts about Seattle’s biggest local brew. Among them: 80 percent of Americans live within 20 miles of a Starbucks. And while you can’t avoid passing a Starbucks in Seattle, you can be a whopping 140 miles away from one in a remote point in Wyoming’s Grand Teton National Park.
Though the project got some attention last fall, he freshened it up for a presentation at the grab-bag speaker series Ignite Seattle (which I emcee) last month. This week he answered some questions about the project.
Q: So why Starbucks?
Starbucks was the gateway for my own love affair with coffee, as it is for so many people. It’s also a Seattle icon, and one of the best ambassadors for our city’s culture the world over. My wife and I have lots of hometown pride.
Q: How did this project come about?
I enjoy maps and coffee, so it made a natural combination. I had done a previous small study on the locations of cafes across the University of Washington’s campus, finding that you’re always within two-minute walking distance (!). I wanted to do the same sort of study across the United States since I had also been playing with Census data for an article on race.
Q: What do you make of the findings? Any of them surprise you?
The map immediately looked great, and got me really excited to dig deeper. At first blush the locations all followed major freeways and cities, and it seemed pretty straightforward. It wasn’t until I started counting the number of people within given distances to Starbucks locations that I was really surprised. I grew up in a pretty small town in Eastern Washington, but only 10 or 15 miles from a Starbucks in the next city.
The punchline of 80 percent of us living within a mere 20 miles was absolutely fascinating to me. While you might not drive 10 or 20 miles to a Starbucks, you definitely will drive that far to buy food or household items, and might stop for a coffee while you’re out! It’s urbanization in its purest form.
Q: In layman’s terms what tools did you use to do the analysis?
The software I used was basic data analysis and visualization tools that we use in astronomy/physics. The location data was saved essentially in a big spreadsheet, which was then mapped by ZIP code. All of this can be done with open source tools, for example Python, which is an exciting part of the data-driven future; anyone can be a part of it.
Q: Have you shared this with anyone at Starbucks?
From my webpage’s hit counter I know that a couple people from Starbucks have read the post, but no formal interactions yet. But I’d absolutely be interested in doing an expanded version of the map in collaboration with them!
Q: How close do YOU live to a Starbucks?
About three-quarters of a mile, almost walking distance. Though, we usually go to a location by the grocery store, which is about two miles.
Read more about Davenport’s methodology and findings on his If We Assume blog.