Analysis of Ben & Jerry’s products, sales, and customers

library(tidyverse)
library(skimr)
library(ggrepel)

ice_cream <- read_csv('https://bcdanl.github.io/data/ben-and-jerry-cleaned.csv')
DT::datatable(ice_cream |> head(100))

DT::datatable(skim(ice_cream))

Money spent on Ben & Jerry’s per flavor

price_house <- ice_cream |> 
  group_by(household_size) |> 
  summarize(tot_spent = sum(priceper1)) |> 
  arrange(desc(tot_spent)) |> 
  mutate(tot_spent = round(tot_spent, 2))

ggplot(data = price_house,
       mapping = aes(x = household_size,
                     y = tot_spent)) +
  geom_line(color = "deepskyblue", size = 1.5) +
  geom_point() +
  theme_minimal() +
  scale_y_continuous(breaks = c(5000,10000,15000,20000,25000,30000), labels = scales :: dollar) +
  scale_x_continuous(breaks = c(1,2,3,4,5,6,7,8,9)) +
  labs(x = "People per Household",
       y = "Total Money Spent",
       title = "Total Money Spent on Ben and Jerry's \n Ice Cream by Household Size") +
  annotate("rect",
           xmin = .75, xmax = 2.25,
           ymin = 17000, ymax = 28500,
           fill = "green3", alpha = .3) +
  annotate("text",
           x = 2.5, y = 25000,
           label = "Households with a small number of \n kids or no kids at all buy much more \n ice cream.",
           hjust = 0, size = 3)

This may be suprising results at first look. One might think that larger households would buy more ice cream since there are most likely more children in the household, and ice cream is notably a favorite treat among children. However, for Ben & Jerry’s that is not the case. Here are a couple reasons why:
1. Smaller portions
Ben and Jerry’s is purchased in either 16 ounce or 32 ounce containers (1 or 2 pints). Purchasing enough Ben & Jerry’s ice cream for a larger household will get very expensive. For large families, it would make more financial sense to get a larger container of ice cream of a different brand for the whole family. On the other hand, a smaller container of ice cream would suit households with 1 to 2, or sometimes 3 people. Buying for a smaller amount of people makes Ben and Jerry’s more convenient, which might lead to a smaller household consistantly buying it.
2. Income
Households with 0 children or 1 child will most likely have different budgetary restrictions that than households with multiple children. Obviously, it is much less expensive to support fewer children. Households with multiple children will have to allocate their money differently, which might deter them from buying multiple containers of Ben and Jerry’s consistantly.

Top 10 Flavors Sold By Region

tot_spent_flavor <- ice_cream |> 
  group_by(flavor_descr, region) |> 
  summarize(flav_spent = sum(priceper1))

ggplot(data = tot_spent_flavor,
       mapping = aes(x = fct_reorder(flavor_descr,
                                     flav_spent,
                                     na.rm = T), 
                     y = flav_spent,
                     fill = flavor_descr))+
  geom_col(data = tot_spent_flavor |> 
             filter(region == "Central") |> 
             arrange(desc(flav_spent)) |> 
             head(5)) +
    geom_col(data = tot_spent_flavor |> 
             filter(region == "East") |> 
             arrange(desc(flav_spent)) |> 
             head(5)) +
    geom_col(data = tot_spent_flavor |> 
             filter(region == "South") |> 
             arrange(desc(flav_spent)) |> 
             head(5)) +
    geom_col(data = tot_spent_flavor |> 
             filter(region == "West") |>
             arrange(desc(flav_spent)) |> 
             head(5)) +
  facet_wrap(~ region, 
             scales = "free_x",
             nrow = 1) +
  theme_bw()+
  theme(legend.position = "bottom",
        legend.key.size = unit(8, "pt"),
        axis.text.x = element_text(angle = 45,
                                   size = 4))+
  scale_fill_viridis_d()+
  scale_y_continuous(labels = scales :: dollar) +
  labs(x = "",
       y = "Total Money Spent",
       title = "Top 5 Flavors Sold by Region",
       fill = "Flavor")