Data Analysis on US Beer Industry
The skills the author demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.
Introduction
When it comes to alcoholic beverages, beer is still the most popular choice. According to a data study by Gallup, the US beer industry has more than $100 billion of sales in 2018. There are over 7,000 breweries currently operating in the US, producing over 100 varieties of beer. Leveraging approximately 10 years of beer review data from beeradvacate.com, I created a Shiny dashboard to help analysts to better understand some of the underlying trends within the beer industry.
Data on Beer Style Concentration
To better analyze the data, I bucketed over 100 beer styles into 10 categories. For example, American Porter and English Porter are both categorized under Porter. Each dot represents a beer style with color indicating its style category.
Beers in the top right quadrant are the ones that have high ratings and high concentration of breweries. Take American India Pale Ale (IPA), there are over 6,000 unique breweries actively produce it in the US. Furthermore, two of the top three concentrated beer types are in the IPA category, which makes IPA the most crowded beer segment.
On the flip side, the bottom left quadrant seems to represent good opportunities because the ratings are low (more room to improve) and there are fewer breweries (less competition). Out of the five beer types, three of them are niche - American Malt Liquor, Japanese Rice Beer, and Low Alcohol Beer.
The other two are American Adjunct Lager and American Light Lager, also known as, the Bud heavy and Bud Light of the world. With fewer than 500 breweries in this segment, it is dominated by beer juggernauts such as Anheuser-Busch InBev, Molson Coors, and Constellation Brands. According to a study done in 2018, all top 10 best-selling beers are American Adjunct and American Light lagers produced by these three companies.[1] Although there is significant potential to improve the quality of the beer, it is difficult to compete with entrenched multi-nationals that spends billions of dollars on advertisement and distribution network.
Do Americans prefer light or heavy?
Since its introduction to the market in the 1980s, American light beer was designed to provide an alternative to those who are health conscious.[2] However, do Americans really like light beer? To answer the question, I compared the average ratings of heavy and light beers from some of the best-selling brands - Budweiser, Coors, Corona, Modelo, Busch, Natural Ice, and Miller. Since the early 2000s, heavy beers have been consistently receiving a higher rating than their light counterparts by a quite sizable margin of 0.4. Light beer is more drinkable and healthier but the benefits do come at the expense of taste.
Data on Alcohol %
In the last section, we learned that beer drinkers like heavy beer more. But where is the sweet spot? For the Pale Lager and Pilsner category, there is a quite strong positive correlation between alcohol percentage and rating in the low alcohol range (5% or below) where most of the beers are. Light beer generally has 4% alcohol content while heavy beer 5%. Once past 5% threshold, the rating increase becomes more gradual as alcohol % increases until it peaks around 7.5%.
Alcohol content % has an impact on the popularity of Pale Lager and Pilsner. However, it is not true for some of the other beer categories. For example, India Pale Ale's rating stays relatively consistent across the alcohol spectrum. For more insights like this, visit my Shiny Dashboard.
Conclusion
Publicly available online beers reviews provide a valuable window into timely consumer preferences and feedback. Understanding the underlying trends is critical for players in the industry to better position themselves in the market place and create better products for the end consumers. If you have any questions or interest to work together on such data, please leave a comment below.
About Author
