top of page
Writer's pictureImon Rashid
Nov 222 min read

Sales Analysis Project Summary

Imon Rashid

November, 2024

All Rights Reserved


This project aimed to analyze weekly sales data from a large retailer to identify key factors influencing sales and develop predictive models for better decision-making. The dataset contained 8,000 records with variables such as store IDs, weekly sales, holiday indicators, temperature, fuel price, CPI, and unemployment rates.


Key Objectives:

  1. To determine how holidays and external factors like temperature and fuel costs influence weekly sales.

  2. To build machine learning models that accurately predict sales trends.


Key Findings:


  1. Holiday Impact:

    • Holidays significantly boosted weekly sales. A "Holiday Impact" metric revealed a strong correlation between holidays and sales performance.

    • Correlation between Holiday_Flag and Weekly_Sales: 0.88.


  2. Economic Factors:

    • Variables like CPI, unemployment, and fuel prices had minimal impact on weekly sales, suggesting these factors are less critical for this retailer’s performance.


  3. Model Performance:


    • Random Forest Regressor outperformed Linear Regression:

      • R² Score: 0.9991 (explaining 99.91% of the variance in sales).

      • Mean Squared Error (MSE): 0.000078 (very low error rate).


  5. Feature Importance:

    • The most influential feature was Holiday_Flag, followed by temperature. Economic indicators like CPI and unemployment ranked lower in importance.


Actionable Recommendations:


  1. Holiday Sales Strategies:

    • Focus marketing efforts and promotions during holidays to maximize revenue.

  2. Ignore Weak Predictors:

    • Factors like CPI and fuel price may not require significant focus, given their weak impact on sales.

  3. Future Data Collection:

    • Collect more granular data, such as daily sales and customer demographics, to uncover additional trends.


Conclusion:



This analysis highlighted the importance of holidays in driving sales and demonstrated the effectiveness of machine learning models like Random Forest for predictive accuracy. By focusing on key drivers like holidays and regional temperature variations, the retailer can optimize sales strategies and improve overall performance.








11 views0 comments

Recent Posts

See All

Comments

bottom of page