☕ ☕️ Cafe Financial Analysis & Insights

Project Overview

This project was conducted to help a small café owner better understand their sales trends. By analyzing transaction-level data, we aimed to identify peak sales times, best-selling items, and customer behavior across the week and year. The insights are intended to guide smarter inventory and marketing decisions.

🧠 Business Problem & Stakeholder

The café owner wanted to answer several key questions:

• What are the peak sales months?
• Are weekends busier than weekdays?
• Which products generate the most revenue?
• How reliable is the dataset? The primary stakeholder is the café owner, who can utilize these insights to enhance profitability and operational efficiency.

🧹 Data Cleaning Choices & Justification

• Data Type Corrections: Converted Transaction Date to datetime; Price Per Unit, Quantity, and Total Spent to numeric
• Missing Data: Used mapping dictionaries and Excel filters to infer missing item names and prices based on other columns
• Duplicates: Checked for and found no duplicate rows
• Dropped/Imputed Columns: Replaced 'ERROR' and 'UNKNOWN' entries with NaN; inferred some missing values via formulas (e.g. Price = Total / Quantity)
• Feature Engineering: Created Month and Day Of The Week columns for further analysis

🔍 Key Findings

• Most sold item by quantity: Sandwich
• Highest revenue item: Salad

• Peak Month:
• Maximum in June (7353.0), minimum in February (6644.0).

• Busiest Day:
• Maximum sales on Thursday $12401.5, minimum - on Wednesday $11680.5. The difference is small, so the demand is stable.

• Weekends showed higher average spending
• Payment Method: Payment method was roughly equally distributed across categories, with each method taking roughly 33% of the total
• Payment method and ordering location do not affect the sales volume.

🤔 Reflections

• Challenges: Ambiguous item-price mapping (e.g. $3 could be Cake or Juice, or $4 could be Smoothie or Sandwich)
• Biases: Imputations were based on assumptions; might not fully represent real-world behaviors
• Next Time: Add a unique transaction ID and explore customer loyalty patterns or time-of-day effects

✅ Takeaways & Recommendations

• Salads and Sandwiches and Smoothie generate the most revenue

• Increase inventory of Sandwiches and Salads

• Focus staffing and marketing efforts on weekends and Thursdays

• Use slow months to experiment with discounts or new menu items

• Potential to increase the average total spending by offering combos

• Consider promotions on days with slightly lower demand (e.g. Wednesday)

• Improve data entry to minimize 'ERROR'/'UNKNOWN' records in the future

  Project Management

  Kudo board

People who collaborated on this project

• Ibrahima Diallo LinkedIn
• Vincent Perez LinkedIn