Introduction

Do household size and use of coupons affect the category of food products purchased by households? This is an important question because marketing of the right coupons to the right households can increase sales of new products.

The way I went about this analysis was by starting with the demographics, transactions, coupon redemption, products data. The four data frames were inner joined by household_id, and product_id. I first joined demographics and coupons to start to find any coupon trends within different demographics. Second, I started to look for patterns in the category of products that were purchased and not purchased by varying household sizes. Lastly, I investigated coupon trends for these products and was able to draw a conclusion from this analysis.

This analysis shows will help the Regork CEO because it pinpointed types of products that we can increase marketing towards to get families to purchase new products. My proposed solution is to increase coupon marketing of longer shelf-life items towards smaller families. The second half of the proposal is to increase coupon marketing of convenient food or meal items for larger families. Large families already purchase high amounts of cold cereal and frozen pizzas, and this is an opportunity to increase exposure to these types of products. It is important to market with coupons and not just the product because households are more inclined to purchasing products that have a coupon.

Packages Required

Packages needed prior to analysis

library(tidyverse) #needed to perform certain functions
library(completejourney) #needed to gain access to data set
library(dplyr) #needed to perform certain functions
library(ggplot2) #needed to create plots

Exploratory Data Analysis

The chart below shows the percentage that each household has saved.

## Joining, by = "household_id"
Household Size Total Spent Total Saved Percent Saved
1 754649.0 132248.38 14.91135
2 1024643.2 178447.07 14.83239
3 383256.2 69130.39 15.28127
4 207407.1 39664.22 16.05376
5+ 255024.3 44806.64 14.94397

The first graph shows the top 10 most purchased items, and the chart is color code by household size. You can see that there are a lot of products that are a part of each household’s top ten but there are a few differences that are split between the smaller and larger family sizes.

## Joining, by = "household_id"
## Joining, by = "product_id"
## Selecting by n

The second graph shows the least purchased items based on household size. There is more variation between the products that households purchase the least and there is an opportunity to increase exposure of these products to increase sales.

## Joining, by = "household_id"
## Joining, by = "product_id"
## Selecting by n

The table below shows the coupon usage for the top ten products that differing sized household’s purchase. The products that households tend to purchase the most, also have the most coupons used.

## Joining, by = "household_id"
## Joining, by = "product_id"
## Joining, by = "household_id"
## Selecting by n
Product Category Number of Times a Coupon Was Used
FLUID MILK PRODUCTS 84487
SOFT DRINKS 81990
BAKED BREAD/BUNS/ROLLS 78884
CHEESE 77249
BAG SNACKS 60380
FRZN MEAT/MEAT DINNERS 55378
YOGURT 51281
SOUP 43996
BEEF 43084
VEGETABLES - SHELF STABLE 42407

The table below shows the coupon usage for the least purchased items depending on household sizes. These products have the least coupon usage compared the products the top ten products.

## Joining, by = "household_id"
## Joining, by = "product_id"
## Joining, by = "household_id"
## Selecting by n
Product Category Number of Times a Coupon Was Used
MISCELLANEOUS CROUTONS 4
NDAIRY/TEAS/JUICE/SOD 4
BULK FOODS 9
COUPON 14
ORGANICS FRUIT & VEGETABLES 14
PROD SUPPLIES 15
SERVICE BEVERAGE 17
NATURAL VITAMINS 25
RESTRICTED DIET 49
MEAT SUPPLIES 51

The two tables above show that there is a high correlation between coupon usage and purchases which is why the new marketing technique will need to include coupons or else the families will not be included to purchase new items.

Summary

The business question addressed was how to market towards different household sizes. It was found that the type of product and coupons make a difference in what different families purchase. The first chart shows how coupons usage differs between household size and it was found that family size does not make a difference in coupon usage. This is valuable information because the use of coupon marketing does not have to differ between families but only what type of coupons are used. By finding the top ten most purchases by smaller families, it was found that they tend to go for products with longer shelf-life where larger families tend to go for products that are convenient to make for meals. The graphs were used to find patterns between size and most and least purchased items. The charts were used to analysis coupon usage, and it was found that families all purchase more items that have coupons verse those that don’t. So that means it is important to prompt coupons to get families to try difference products and to tailor these coupons to the type of products that each families tend to get.

I propose to increase coupon marketing of long shelf-life items to smaller families and to increase coupon marketing of quick, easy food items to larger families. There are some limitations to this data because I did not get into the specifications of the household that may have young children which could cause different eating habits than families with all adults. It will be important to start to factor in other aspects of the households besides just family size.