This project analyzes promotional campaign performance for an FMCG client using sales and events data. It focuses on understanding how promotions impacted revenue, sales volume, stores, cities, and product categories.
The goal is to turn raw business questions into clear insights that help teams evaluate campaign success, identify top-performing promotions, and improve future promotional planning.
The project was built to answer practical business questions such as:
- ✅ Are there duplicate records in the events data that need to be removed?
- 🏬 Which cities and stores contribute the most to sales performance?
- 🧹 How should missing sales values be handled for reliable analysis?
- 🎉 Which campaigns performed better after promotions?
- 📈 Which products, stores, and promo types delivered the strongest uplift?
⚠️ Where did promotions underperform in terms of revenue or sold units?
The analysis covers both data preparation and business insight generation.
- Removes duplicate event records based on key business identifiers
- Handles missing values in sales-related columns
- Prepares clean data for campaign and promotion analysis
- Compares performance before and after promotions
- Evaluates campaign effectiveness for Diwali and Sankranti
- Measures uplift using Incremental Revenue Percentage (IR%) and Incremental Sold Units Percentage (ISU%)
- Identifies high- and low-performing products, stores, cities, and promotion types
- Builds visualizations to communicate business insights clearly
IR% measures how much revenue changed after a promotion compared to before the promotion. It helps show whether a campaign improved revenue performance.
ISU% measures how much unit sales changed after a promotion compared to before the promotion. It helps show whether a promotion increased product movement.
This project includes analysis such as:
- 🧾 Duplicate row detection and removal
- 🏙️ Store count analysis across cities
- 📌 Missing value imputation using median quantity sold before promotion
- 🏷️ Product category pricing comparisons
- 🎁 BOGOF performance during the Diwali campaign
- 🏪 Store-level sales performance during major campaigns
- 🪔 Campaign comparison between Sankranti and Diwali
- 📊 Product-level and store-level uplift analysis using IR% and ISU%
- 🚫 Identification of negative-performing promotion types
This chart shows the number of stores present in each city.
This visualization highlights the presence of distinct store chains across cities.
This chart shows how different product categories contributed to overall quantity sold after promotion during the Sankranti campaign.
This visualization examines the relationship between post-promotion base price and quantity sold after promotion.
These distributions help compare how quantity sold before promotions varies across product categories.
This chart compares Incremental Sold Units Percentage across different cities to understand promotion effectiveness by location.
This scatter plot helps compare incremental revenue and incremental sold units across different promotion types in Hyderabad.
This chart compares category-level revenue before and after promotions in Bengaluru.
This project helps business teams:
- ✅ Improve data quality before analysis
- 📈 Measure the real impact of promotions
- 🎯 Understand which campaigns and promo types work best
- 🏪 Compare regional and store-level performance
- 💡 Support better planning for future promotions using data-backed insights
The presentation deck for this project can be found here:
Artifacts/presentation_deck.pptx
The datasets used in this project have not been shared in this repository due to client confidentiality reasons.
This repository presents a simple business-focused analysis of FMCG promotion performance. It combines data cleaning, campaign evaluation, KPI measurement, and visualization to help stakeholders understand how promotions influenced revenue and sales across products, stores, and cities.
This project is licensed under the MIT License.