Learn practical skills, build real-world projects, and advance your career

Business and Operation Analysis of a New Store

Background

A Store has opened roughly for 5 weeks. Unfortunately, the traffic system has not gone live yet as it is still under going tuning. However, the operations team is eager to study the staff allocation based on the most recent 4 weeks of business. Three areas are identified to look into with the absence of the traffic data:

  1. How is the business in this new store without the sales of hot-selling items?
  2. When is a good lunch break period for staff?
  3. Does the store need more staff?
  4. How is the business doing in the Kids' corner?
  5. Are there enough staff servicing the consumers in the Kids' corner?
  6. What suggestions for improvement can be made from the study?

Due to the lack of traffic data, the transactional data is used as an alternative for the research. From IT departement, extracts of transactional data and the product reference data are collected.

Data Sets

Transactional data (stored as TrxData.csv) shows all the transactions occurred entirely in the course of 4 weeks. It consists of the following attributes:

  • Receipt- this is to track each and every purchase occurred.
  • Timestamp - this is to track when the purchase taken place.
  • Item - what is being purchased.
  • Qty - the quantity sold.

Product data (stored as ProdDef.csv) is just a product reference data set. It assists for one to better knowing the product's different classifications. It consists of the following attributes:

  • Item: what is being purchased.
  • C1: this is the classification on gender and age.
    • In gender, there are 3 categories:Male, Female and Unisex.
    • In age, there are 4 categories: Adult(16+), Grade School(12-16), Pre-School(4-11), Toddler(0-3).
  • C2: this is to classify the products into Sports and Leisure.
  • C3: this is another product classification among Shoes, Clothing and Equipment.
  • Marketing - an inductor to show high-demand product.

Looking at these 2 data sets, the TrxData.csv file is almost good to use with decomposing the timestamp into useful date/time intervals, while the ProdDef.csv file may need some transformations to better serve our purpose - unemployed attributes will be removed and kids' products need to be identified.

Without further ado, let's first upload the Jupyter notebook (new-store-analysis.ipynb) to Jovian.ai with the 2 above mentioned data sets: TrxData.csv and ProdDef.csv.

project_name = "new-store-analysis"
!pip install jovian --upgrade -q
import jovian