What Can We Learn From Starbucks?

For some people, their morning cup of coffee is crucial to survival for the day. However, the way in which people obtain this cup of coffee differs on many factors like; if they have time to make it themself, if they have the machines to make the coffee, if they have coffee beans, or if they are just going to pay for someone else to do the work for them. As a broke college kid myself, I do not have the extra cash laying around to be able to pay for someone to make me a delicious cup of coffee every morning (so I make it work by using my roommates Keurig and buying massive Costco boxes of the K-cups and hoping for the best). Because I cannot regularly pay for someone else to make me a special coffee, I view a Starbucks coffee as a luxury item. Although I cannot regularly afford Starbucks, I hypothesize that people with a more flexible income and financial comfort could afford items like a Starbucks coffee more regularly. I also speculate that when Starbucks franchisee owners are looking to open a Starbucks location, they do their research to theorize where concentrations of wealth are, meaning areas where people can afford to buy Starbucks regularly. To spin their research on its head, I was curious if I could determine concentrations of wealth in the US by first determining concentrations of Starbucks locations. I wanted to see if there was a correlation between Starbucks locations, and high per capita income.

I first got information about Starbucks locations by using Chris Meller’s StarbucksScraper, it pulls data from the Starbucks Store Locator API, which contains locations of Starbucks all over the world. I imported the gathered csv into a Jupyter notebook and got to work, first just looking at the beginning and end of the data file to get a look at what I’m working with and make sure its properly formatted.

I decided to just focus on Starbucks locations within the United States for the sake of relevance. I then also fact checked to make sure that the list included all 50 states (I found that it included 51, but by reading over the list I saw that it counted DC as its own state).

Then I wanted to visualize where these Starbucks locations are, so I imported it to Tableau and created a geographical representation of the data which helped to get a better grasp of their location.

I could first establish that there was a large concentration of Starbucks in California, which I thought was very interesting.

Next I wanted to drill down the information and look at specific city locations of the Starbucks.

This second visual used the exact same data, however, it brought light more specific concentrations of locations (like how in California we can see there is a concentration of locations near the coast and San Francisco area). There is also a concentration of Starbucks in the North West corner of Washington, I hypothesize that this is because Seattle is where the original Starbucks first was, and thus increasing the popularity of Starbucks.

Now that I had a grasp of where Starbucks are most commonly located, I turned to the United States Census Bureau to gather information about peoples per capita income. I decided to use per capita income because I thought it would get a clear average of each persons yearly income, I also gathered data from 2017 so it would match the Starbucks timeframe.

I first sorted the per capita data by city, but because they are so small and concentrated it was hard to see much discrepancy (besides a wealthy cluster on the east coast).

So I also represented the data by state, which indicated a concentration in Texas, Virginia, and Georgia.

At this point I started to see some holes in my research, and thought about what other factors may affect Starbucks locations. So I decided to also gather information about population to cross check my findings and try to calibrate for areas of lower populations (I used population data from the Census Bureau).

One thing that was interesting was that although California has the highest population, it is ranked 22nd for per capita income.

After looking at how each state ranks in the three separate categories (Starbucks locations, per capita income, and population) it seams as though there is not a strong enough correlation to be able to confidently identify concentrations of wealth by locations of Starbucks. For example, by looking at just California’s rankings (high in population and Starbucks locations but low in per capita income), it seams as though there is not much correlation. These findings, however, have lead me to hypothesize what other reasons could there be for these concentrations of Starbucks? One parameter I had not considered earlier is also the persona that the brand Starbucks carries with it. Meaning that different cultures in different states may have a larger impact on Starbucks locations than concentrations of wealth.

Further research questions I would like to investigate would be if there is possibly a different restaurant that may be a better indicator of a states financial status. For example, maybe there is a correlation between McDonalds locations in areas of poverty because of their target market and business plan in being a low cost fast-food restaurant.

This research also pushed me to question my preconceived notion about how people spend their money and that maybe my idea of Starbucks is false. It reminded me to try and think with a broader perspective and that sometimes the things I think are normal and just a part of everyday life, may be weird and abstract to others.