Final Project

Introduction

Regork grocery chain caters to a wide demographic. Based on the data, it would be best to change that strategy so that the optimizations can improve our customer satisfaction and financial metrics. We have analyzed the most important customer demographics based on income bracket. I propose we optimize various departments of this particular store to cater to the main customer demographic.

Packages Used

tidyverse - a collection of packages for data wrangling, exploration and visualization

completejourney - data set for Regork grocery store

Lubridate - allows date/time calculations

ggplot2 - visualization tool for graphs and charts

library(completejourney)
library(tidyverse)
library(lubridate)
library(ggplot2)
library(dplyr)

Data Preparation/Exploration

Sorting total sales by customer income

Preparing dataset

Prod <- products
tran <- get_transactions()
demo <- demographics

X1 <- inner_join(tran, Prod, by = "product_id") 
D1 <- inner_join(X1, demo, by = "household_id") 

Analyzing Data

D2 <- D1 %>%
  group_by(income) %>%
  summarise(total_sales = sum(sales_value, na.rm = TRUE)) %>%
  arrange(desc(total_sales)) %>% 
  mutate(SUM = sum(total_sales), Percent = (total_sales/SUM)*100)

Visualize Data

Finding the most popular week days for our target demopraphic ($1k-99k income)

Preparing dataset

D1$weekday <- wday(D1$transaction_timestamp)
D1$hour <- hour(D1$transaction_timestamp)
target_demo <- c("50-74K","35-49K", "75-99K")

Analyzing Data

D3 <-  D1 %>%
  group_by(weekday) %>%
  filter(income %in% target_demo) %>%
  summarise(total_sales = sum(sales_value, na.rm = TRUE)) %>%
  arrange(desc(total_sales)) %>%
  mutate(SUM = sum(total_sales), Percent = (total_sales/SUM)*100)

D3$weekday[D3$weekday==1]<-"SUNDAY"
D3$weekday[D3$weekday==2]<-"MONDAY"
D3$weekday[D3$weekday==3]<-"TUESDAY"
D3$weekday[D3$weekday==4]<-"WEDNESDAY"
D3$weekday[D3$weekday==5]<-"THURSDAY"
D3$weekday[D3$weekday==6]<-"FRIDAY"
D3$weekday[D3$weekday==7]<-"SATURDAY"

Visualize Data

Finding the most popular hours for our target demograhpic ($1k-$99k)

Analyzing Data

D4 <-  D1 %>%
  group_by(hour) %>%
  filter(income %in% target_demo) %>%
  summarise(total_sales = sum(sales_value, na.rm = TRUE)) %>%
  arrange(desc(total_sales)) %>%
  mutate(SUM = sum(total_sales), Percent = (total_sales/SUM)*100) %>%
  slice(1:10)

D4$hour[D4$hour==17]<-"5 PM"
D4$hour[D4$hour==18]<-"6 PM"
D4$hour[D4$hour==16]<-"4 PM"
D4$hour[D4$hour==19]<-"7 PM"
D4$hour[D4$hour==13]<-"1 PM"
D4$hour[D4$hour==15]<-"3 PM"
D4$hour[D4$hour==14]<-"2 PM"
D4$hour[D4$hour==12]<-"12 PM"
D4$hour[D4$hour==20]<-"8 PM"
D4$hour[D4$hour==11]<-"11 AM"

Visualize Data

Finding the top 10 most popular products for our customer demographic

Analyzing Data

D5 <-  D1 %>%
  group_by(product_category) %>%
  filter(income %in% target_demo) %>%
  filter(product_category!="COUPON/MISC ITEMS") %>%
  summarise(total_sales = sum(sales_value, na.rm = TRUE)) %>%
  arrange(desc(total_sales)) %>%
  slice(1:10)

Visualizing Data

Summary

The problem identified through exploratory analysis was that this particular store was not catering it’s main customer demographic. Customers making $35K-$99k make up 55% of total sales. Customers making less than $35k made up 21% of total sales. Customers making more than $99k made up 26% of total sales. The solution is to optimize inventory, schedule, and other attributes of the store to focus on low-middle class customers. From now on the target demographic are customers making $1k-99k. Our strategy should focus on conforming our product types to this income bracket with decent packaging to appeal to higher income brackets. We need to reevaluate the products in soft drinks, beef, milk, cheese, frozen dinners, snacks, bread, beer, frozen pizza and deli meats and make sure that it appeals to our target income bracket. And optimize our staff scheduling/supply chain management to account for the most common times our target demographic shops which is - Sunday, Monday, Friday and Saturday from 4PM - 7PM. We could improve this analysis by obtaining more attributes regarding product inventory and customer demographics.