Packages

Question

As a senior in college, I am looking forward to graduation and celebrating with my family and best friends. When I’m celebrating, there’s nothing better than an ice cold alcoholic beverage. And these days who doesn’t love a good Whiteclaw? Whiteclaws took the United States by storm in the past few years especially in the college scene. Large household names like Bud Light, Barstool Sports, Carona and more have tried to recreate their own version of the Whiteclaw and it has led to some serious debate. I want to find out which people are more satisfied drinkers with Whiteclaws or traditional beer brands. Are consumers more satisfied with Bud Light or are they more satisfied with the new to the scene Whiteclaw? I want to find out because the brand that has the most satisfied consumers is the one I will purchase for my graduation celebration with my friends. I am looking forward to sharing my findings with you.

Data Collection

Im planning on collecting the data from Twitters API. I will be able to search for the twitter handle of Bud Light and Whiteclaw with bouillon logic from the searh_tweets2 functions. I can also use that logic to seach for the times where users tweet about the two brands. This will provide me with many rows of data that I can compare back and forth to determine consumer sentiments. I am assuming that Bud Light will come back with more observations as it is a larger company that has been around a lot longer and they have more followers on Twitters platform.

Set up Rtweet Environment

To set up our environment we need to use the key, token, and app that allows us access like an account to the information.

Twitter API Data

We are going to search for different words that give us data on whiteclaw and bud light to discover sentiments. After that we are going to convert the API data into 2 separate csv’s. Then we will bring back the csv’s for further analysis.

Next we will set up the script for sentiments. This is using the package that is built into R studio.

nrc <- get_sentiments("nrc")
bing <- get_sentiments("bing")

3 - Mutating, Aggregating, Cleaning Data

At this point we have loaded the data into R Studio as a CSV file and will need to modify it for further analysis. We will be aggregating the data by combining certain columns and identifying sentiments from each tweet later on. We will perform mutate functions to specify which product is being referred to in each row - Whiteclaw or Bud Light.

## # A tibble: 2 x 2
##   Beverage      n
##   <chr>     <int>
## 1 Bud Light  2697
## 2 Whiteclaw  1015
## # A tibble: 1,608 x 3
## # Groups:   Beverage [2]
##    Beverage  word        n
##    <chr>     <chr>   <int>
##  1 Bud Light bud       202
##  2 Bud Light light     196
##  3 Bud Light https     130
##  4 Bud Light t.co      130
##  5 Bud Light sports     65
##  6 Bud Light summer     65
##  7 Bud Light tickets    63
##  8 Whiteclaw white      61
##  9 Bud Light live       58
## 10 Whiteclaw claw       58
## # ... with 1,598 more rows

4 - Performing the Analysis and Communicating the Results

Visual 1

Comparing Sentiments of Bud Light and Whiteclaws

I wanted to perform an analysis that compares the sentiment between the two beverage companies - Whiteclaw and Bud Light. On the X axis the sentiments are displayed and on the Y axis it displays the number of words that were used that fall under the each sentiment. In the blue, we can see the results for Whiteclaw tweets. In the red, we can see the results for Bud Light Tweets. One thing to keep in mind is that Bud Light has roughly 200 total tweets compared to Whiteclaw’s 65 total tweets.

Through this visualization, we can see some mixed results but we can determine a few things. On one hand, Whiteclaw has far less reviews than Bud Light in our data, however, they are much more often associated with the sentiments of anger, negativity, and fear. On the other hand, Whiteclaw is also associated with trust and anticipation more than Bud Light is.

I can attempt to guess at why this may be. Perhaps the hard seltzer company has not been making new flavors or varieties, which makes their customers/fans angry. But maybe, their fans are also anticipating a new flavor or variety to come out and they trust in Whiteclaw to keep them on their toes with delicious new types of hard seltzers.

Bud Light has higher numbers in sentiment categories such as Joy, positive, and still have high numbers for anticipation! Their positive tweets are enormously higher than Whiteclaw’s and their fear is almost half of Whiteclaw’s. Bud Light, owned by Anheuser Busch, has a larger following and has been a household name for nearly 50 years. This might help explain some of the positive sentiment surrounding this brand of beer.

New data frame of Tweet Sentiments was created at this point

A new table was made adding the tweet sentiments to the original combined data of Whiteclaws and Bud Light. From there I will be able to make more summarizations with the sentiments of the tweets.

Visual 2 - Summary Table

Creating a Summary of Positive and Negative Sentiments for Whiteclaw and Bud Light
## # A tibble: 2 x 8
##   Beverage `sum(positive)` `sum(negative)` `sum(disgust)` `sum(joy)`
##   <chr>              <dbl>           <dbl>          <dbl>      <dbl>
## 1 Bud Lig~             190              74             27        128
## 2 Whitecl~             137              89              5         92
## # ... with 3 more variables: `sum(anger)` <dbl>, `sum(sadness)` <dbl>,
## #   `sum(positivity)` <dbl>

Above you will see the sum of words that are associated with different sentiments. These are words that I believe are good indicators of how what the tweets were meant to portray. Furthermore, this is a good summary that actually allows me to quantify the words associated with sentiments instead of just seeing a visual for the different breakout. In my findings it is strange that positivity and positive do not have the same quantity, because I would believe those two are interconnected. I also found it strange that so many tweets about Bud Light are associated with sadness and positivity because these are polar opposite emotions.

Visual 3

Creating a bar chart for negative and positive during the school months

College kids love to drink bud light and Whiteclaws on the weekends. They are light drinks that taste great in my opinion. So I want to see which brand had the most positive and negative reviews during the last few months at most colleges.

Visual 4

This chart is an interesting analysis that shows some of the words between both the bud light beverage and the Whiteclaw beverage twitters. It provides an interesting analysis for both companies to see when people are tweeting about using their products, or when they are looking forward to using their products. As you can see some popular words on there are: summer, music, sports, endorse. So it’s reasonable to assume that people enjoy drinking during sporting events, or when they are listening to music, and most of all looking forward to having a cold one in the summer months. Surprisingly, one of the words that is not one of the most common is college. I was assuming that college kids would be tweeting about drinking these beverages during school.

Coclusion

In conclusion I think the good old fashioned word cloud sets it up perfectly for me. I ran a sentiment analysis comparing Whiteclaw and Bud Light all in one graph which showed some good comparisons. I realized that Bud Light has positive and joyful emotions attached to their brand. But unlike Bud Light, Whiteclaw had some negative emotions in the tweets abou their brand.

Customers and fans of the popular hard seltzer are anticipating and trusting that Whiteclaw is going to be coming out with new flavors or varieties that they can drink with their friends.

Since I only have a few days left before my graduation celebrations, I will have to choose bud light to be served to my guests. I can’t wait and anticipate any longer for Whiteclaw to come out with a new flavor that brings people positivity. I think when I hand out the beers to everyone I will see so much joy in their faces. But either way, summer is here, baseball is here, so why not sit back and enjoy a drink!