Starbucks store location analysis

Avatar
Posted on Oct 25, 2020

Starbucks is one of the biggest coffee brands in the world. Since founded in 1971 in Seattle, Washington, it has been constantly growing and now has more than 22,519 stores worldwide (starbuscks.com). The purpose of this project is to understand the underlying logic of Starbucks store opening strategy, so as to extract transferable insights for other similar brands.

The primary data set is from Kaggle.com, containing 25,600 Starbucks store location information in 73 countries and areas. The data set was collected in February, 2017. Four additional data sets are also included for a complete analysis, including world country GDP, population, corruption index and country code.  The shiny app is here.

Store Location Overview

The Starbucks store location data set includes 25,600 rows of data, each indicates a unique store information. All these stores are located in 73 countries in six continents, which are North America, South America, Europe, Asia, Africa and Oceania. In United State alone, there are 13608 stores running in year 2017, constitutes over 53% of the total store counts, making North America the top 1 market for Starbucks brand.

 

 

As seen from the bar plot of top 10 country by store count, 5 out of the top 10 country are located in Asia. If we check the top 10 city by store count, the top 3 are also Asian cities. In total, there are 7160 stores existing in Asia, and that make it the second largest market for Starbucks.

 

However, using absolute store count can be influenced by the country size and population. Therefore, a store density metric may be more accurate when measuring how heavily Starbucks investigate in certain area. In this analysis, the number of population served by 1 store (population in thousand/store number) is calculated and used. The result turns out that only 4 of the Top 10 countries have high absolute store counts. The other are all high population density countries or tourist destination.

Store Location Analysis

When considering whether or not to enter a foreign market and make heavy investment in business, one thing every company must consider is the economic status of the local country. Therefore, GDP and population data are something need to look at. When measuring the difference between GDP, population, and GDP Per Capita, countries that have Starbucks stores and those don’t show distinction in all three metrics.  Basically, the average GDP, population, and GDP Per Capita of those countries don’t have a Starbucks stores are all lower than those counties has at least one Starbucks stores, indicating that Starbucks store favors those areas of better economic status.

Note: Outliers are excluded in the GDP graph (USA and China), and in the population graph (China and India)

Even though, there are quite a few outliers exist in the country group that has no Starbucks stores: these countries have a GDP Per Capita higher than that of those counties have Starbucks stores, but there is no store opened in the area. To take a closer look at the outliers, I pulled the names of those outlier countries and did a manual search on Google. It turned out that by 2020, 3 years after the data was collected, 7 out of the 14 countries now have at least 1 Starbucks. Starbucks is not really neglecting these countries, they still try to enter all the countries who are economically superior, only it takes time.

The next step I’m trying to find a quantitative relationship between the store number and the economic indicators. Again, the store density metric is used instead of the absolute store count, so as to eliminate the influence on country size and population.

 

After taking log of both variables, the number of population served by 1 store and GDP Per Capita show an apparent negative correlation relationship in the scatter plot. As country’s GDP Per Capita increases, the average population served by store decreases, meaning in wealthier counties, Starbucks tends to open up more stores there.

 

When color-coded the scatter plot and add an extra dimension of continent information, another interesting pattern can be spotted from the plot. The Asian countries are mostly clustered below the regression line, whereas the European countries all sit above the line. This indicates that in countries of similar GDP Per Capita, Starbucks tends to open more store in Asia, and less store in Europe, after controlling the effect of population.

One possible explanation might be the coffee drinking history and culture in the two continents. In Asia, coffee is a relatively new beverage and people traditionally don’t drink it, so it is easier for Starbucks to enter the market and earn more share with less effort. However, in Europe, many countries already have a history and culture of coffee drinking, which makes it harder for Starbucks to join the competition.

Going back to the outlier countries in the boxplot of GDP, population and GDP Per Capita, Italy is the only country that has higher economic performance than the average of those countries already have a Starbucks, however the first Starbucks store opened after year 2017. It is probably because the coffee culture in Italy is very profound, so Starbucks need to be fully prepared before opening a store there.

Store Ownership Type and Expansion Strategy

There are 4 types of ownership among the 25,600 individual Starbucks stores in the world, except for Company Owned, there are also Franchise, Joint Venture and Licensed. From company perspective, it must be favored to have everything owned by the company alone, so they can have all the revenue and make it easier for the business operation – unless there is some benefits or obligation with other type of ownership.  One possible influence could be the country corruption index, which reflect how easy to do business in a certain country. But when looking at the average index alone at continent level, it is hard for us to notice any pattern.

 

When looking at the store map of United Kingdom, an interesting pattern can be spotted easily. Every dot in the map represent an individual store, the dark green color represent company owned stores, black is franchise stores and light green is licensed stores. In the map, the dark green dots of those company owned stores are mostly clustering in London city, while the black and light green dots are more spread out over the country. This suggests there might be different expansion logics associate with different types of ownership.  The opening of company owned stores depends on the Starbucks company --- after making adequate market research, the company launches a new store in a new area. It is more naturally they will focus on the big metropolitan areas, since population density are higher so as the household income. But in the vast majority of the world, Starbucks offer this opportunity for anyone or companies who think they see a chance of making profit from a Starbucks store to apply for a partnership. Therefore, they don’t need to spend too much energy on researching but always be able to open more stores.

This dynamic may vary from place to place. Conclusions need more data support and analysis.

Average Distance Between Stores in City

On city level, store number may not be an accurate indicator since the area of each city in the world can vary drastically. To measure the store density again, I calculated the average distance between any stores in a city where exist more than 15 stores. It turns out that New York City in the city with the highest Starbucks store density in the world, on average for every 2.3 miles you walk, you will see another Starbucks shops.

 

Despite all other factors, this metric may help us to spot the city with a potential to open more stores: if any city or local neighborhood have a similar coffee-drinking population density, this average store distance may be treated as a growth goal for Starbucks.

Conclusion:

  1. Starbucks opens stores in countries who have a relatively high GDP Per Capita.
  2. In countries of similar GDP Per Capita, Starbucks opens more store in the emerging Asian market.
  3. There is a set of distinct expansion strategies associated with stores of different ownership types, which requires further investigation.

 

About Author

Avatar

Chun Tao

M.S of Education from University of Pennsylvania
View all posts by Chun Tao >

Leave a Comment

No comments found.

View Posts by Categories


Our Recent Popular Posts


View Posts by Tags

#python #trainwithnycdsa 2019 airbnb Alex Baransky alumni Alumni Interview Alumni Reviews Alumni Spotlight alumni story Alumnus API Application artist aws beautiful soup Best Bootcamp Best Data Science 2019 Best Data Science Bootcamp Best Data Science Bootcamp 2020 Best Ranked Big Data Book Launch Book-Signing bootcamp Bootcamp Alumni Bootcamp Prep Bundles California Cancer Research capstone Career Career Day citibike clustering Coding Course Demo Course Report D3.js data Data Analyst data science Data Science Academy Data Science Bootcamp Data science jobs Data Science Reviews Data Scientist Data Scientist Jobs data visualization Deep Learning Demo Day Discount dplyr employer networking feature engineering Finance Financial Data Science Flask gbm Get Hired ggplot2 googleVis Hadoop higgs boson Hiring hiring partner events Hiring Partners Industry Experts Instructor Blog Instructor Interview Job Job Placement Jobs Jon Krohn JP Morgan Chase Kaggle Kickstarter lasso regression Lead Data Scienctist Lead Data Scientist leaflet linear regression Logistic Regression machine learning Maps matplotlib Medical Research Meet the team meetup Networking neural network Neural networks New Courses nlp NYC NYC Data Science nyc data science academy NYC Open Data NYCDSA NYCDSA Alumni Online Online Bootcamp Online Training Open Data painter pandas Part-time Portfolio Development prediction Prework Programming PwC python python machine learning python scrapy python web scraping python webscraping Python Workshop R R language R Programming R Shiny r studio R Visualization R Workshop R-bloggers random forest Ranking recommendation recommendation system regression Remote remote data science bootcamp Scrapy scrapy visualization seaborn Selenium sentiment analysis Shiny Shiny Dashboard Spark Special Special Summer Sports statistics streaming Student Interview Student Showcase SVM Switchup Tableau team TensorFlow Testimonial tf-idf Top Data Science Bootcamp twitter visualization web scraping Weekend Course What to expect word cloud word2vec XGBoost yelp