As my first project using the Tableau visualization software, I plotted every single Starbucks location worldwide to produce this map.
Based on my personal experience, I hypothesised that the density of Starbucks coffee stores would be correlated with income levels of the same region. As Starbucks’ coffee is generally more expensive than generic substitutes, the company would likely only open new stores in locations with sufficient income to sustain the coffee chain.
I downloaded a data set about Starbucks locations from Socrata as a CSV and then loaded them as source data into Tableau. From there, it was a relatively simple matter to use the latitude and longitude entries to create a new sheet and visualize the data.
Comprehensive worldwide data on income levels was difficult to find, as no publisher releases a complete data set on income on a locality basis. Most data sources, such as the World Bank, only provides income information on a national-level resolution. Because I wanted to look into how Starbucks locations are affected by local income variations, I did not use of the existing data sets.
I considered producing a geographic visualization of only the United States, as locality-level income information is available from the United States Census Bureau.
However, I was unable to map the correlation as there was no easy way to sort each Starbucks store into geographic groups. Political boundaries (state, city, village lines) would not provide the right zoning as US economic areas usually transcend political borders. USPS ZIP Codes do not match the underlying geoeconomic activities, as IRS income data is only available based on tax filers’ residential addresses, but workers usually work (and likely consume coffee) in another ZIP Code (e.g. living in a suburb and commuting to work in a downtown business district).
I continue to search for a meaningful data set to correlate Starbucks locations with income levels in the geographic region.