Retail Sales Analysis

A comprehensive analysis of digitized sales data for an independent retail store.

Background

An elderly couple operating a brick‑and‑mortar retail store in Swat, Pakistan, maintained paper records of every transaction. To optimize inventory management, identify best-selling products, and gain a better understanding of customer spending patterns, these records were digitized and analyzed using Python and the pandas library.

Objectives

1. Import & preprocess the raw CSV sales data into a pandas DataFrame.
2. Explore the dataset to understand data types, detect outliers, and inspect top transactions.
3. Aggregate & summarize sales by product and by invoice to identify high‑value purchases and customer spending behavior.
4. Generate descriptive statistics on invoice totals and investigate the relationship between units sold and revenue.
5. Provide actionable insights and recommendations to the store owners.

Data

• Source:

  • data/sales.csv (raw transaction log provided by store owners, uploaded to this repository’s data/ folder)

• Description:

  Column	Type	Description
  invoice_id	int64	Unique identifier for each sale; repeats for multi‑item invoices
  date	object	Date of the transaction (YYYY‑MM‑DD)
  day_of_week	object	Day name (Monday–Saturday)
  product_name	object	Description of the item sold
  units_sold	int64	Quantity of the item sold in that line
  price_per_unit	float64	Unit price charged
  amount	float64	Line total (units_sold × price_per_unit)

Environment

1. Clone the repository

   git clone https://github.com/shahab-ghafoor/retail‑sales‑analysis.git
   cd retail‑sales‑analysis
   
   

Key Findings

• Highest single‑line sale: A vintage kitchen cabinet sold for $ 295.00 in one invoice.
• Top‑selling product by frequency: “HAND WARMER BIRD DESIGN” appeared on 8 invoices, with 14 units sold in total.
• Customer spending behavior (invoices 21–60):
  • Mean invoice total: $ 52.55
  • Maximum invoice total: $ 295.00
  • Total revenue: $ 2,102.03
• Units‑to‑revenue relationship: Moderate positive correlation (r ≈ 0.62), indicating larger orders tend to generate proportionally higher revenue.

Recommendations

1. Stock up on hand warmers and 3‑tier cake tins—top revenue generators.
2. Bundle promotions around frequently co‑purchased items (e.g., hand warmers + seasonal décor).
3. Monitor low‑cost items for volume trends to decide whether to expand or phase out offerings.