| title | Retail Sales Analytics: Comprehensive Data Analysis |
|---|---|
| description | Detailed overview of 10 in-depth retail sales analyses, covering churn, forecasting, segmentation, and more using sales data in US. |
| layout | default |
This page serves as a comprehensive overview of the Retail Sales Analytics project, which consists of 10 in-depth analyses conducted on retail sales data. Each analysis focuses on different aspects of customer behavior, sales trends, and business performance, providing actionable insights for the FMCG and retail industry. These analyses combine various data science techniques, from predictive modeling and clustering to time-series forecasting and geospatial visualization, to support data-driven decision-making.
Key Analysis: The analysis of Average Churn Status by Days Since First Purchase reveals significant variability in churn rates, particularly within the first 2000 days of a customer's lifecycle. Newer customers show a higher propensity to churn, emphasizing the critical need for targeted retention strategies during this early period. As customers mature, their churn rate stabilizes, indicating more consistent behavior among long-term patrons.
Tools Used:
- Python: Data analysis and preprocessing.
- Pandas: Data manipulation and aggregation.
- Matplotlib & Seaborn: Visualization of churn trends.
- Logistic Regression, Random Forest: Predictive modeling for churn.
Conclusion: To effectively reduce churn, focus efforts on engaging new customers early in their lifecycle. Simultaneously, maintain consistent service quality for long-term customers to ensure their continued retention.
Key Analysis: The Sales Monthly Trends in the US show a prominent spike in December, indicating a strong surge in customer purchases during the holiday season. This trend underscores the importance of proactive inventory and logistics planning in the months leading up to year-end. Conversely, sales experience a sharp decline in January and February post-holiday, followed by a gradual recovery in subsequent months. Utilizing ARIMA models, we projected future sales increases for certain months, indicating potential revenue growth.
Tools Used:
- Python: Data analysis and preprocessing.
- Pandas: Data manipulation and aggregation.
- Matplotlib & Seaborn: Visualization of sales and order trends.
- Time Series Analysis, ARIMA: Demand forecasting optimization.
Conclusion: Businesses should strategically plan promotions and optimize inventory for the holiday season while developing strategies to sustain sales during the quieter post-holiday months. Improve stock management to meet predicted demand and capitalize on growth opportunities.
Key Analysis: Customer Segmentation Based on Age, Gender, and Quantity of Orders reveals distinct customer groups. Male customers, especially in older age groups, predominantly fall into Segment 2 (yellow), suggesting this segment might represent a group with higher purchasing power or specific buying behaviors. The Distribution of Products Across Customer Segments shows that "Superstore" and "School & Education" categories dominate order quantities, particularly in Segment 0 (blue), indicating the highest overall demand for these product types.
Tools Used:
- Python: Data analysis and preprocessing.
- Pandas: Data manipulation and aggregation.
- Matplotlib & Seaborn: Visualization of customer segments and product distribution.
- Scikit-learn, KMeans Clustering: Customer segmentation.
Conclusion: Implement personalized marketing strategies tailored to each segment. For instance, target male customers in Segment 2 with premium products, while focusing on broader offerings for females in Segment 0. These insights also suggest prioritizing promotions in high-demand categories while exploring strategies to boost engagement in underperforming categories.
Key Analysis: The Price Elasticity Analysis for Optimizing Pricing Decisions illustrates the relationship between product price and sales volume. Historical sales data (blue dots) informed a regression analysis (red line) which indicates that sales decrease by 0.8 units for every $1 increase in price. For example, at a price of $20, estimated sales are around 86 units.
Tools Used:
- Python: Data analysis and preprocessing.
- Pandas: Data manipulation and aggregation.
- Matplotlib & Seaborn: Visualization of price-sales relationship.
- Regression Analysis: Modeling price sensitivity.
Conclusion: To maximize revenue, businesses must consider the optimal balance between sales volume and price. This analysis provides a data-driven approach to determine pricing strategies that not only consider sales volume but also profit margins, optimizing revenue potential.
Key Analysis: The Purchase Volume by Age Group and Gender analysis shows that the 46-55 age group (category D) is the most active in purchasing products. Across all age groups, men consistently dominate purchases.
Tools Used:
- Python: Data analysis and preprocessing.
- Pandas: Data manipulation and aggregation.
- Matplotlib & Seaborn: Visualization of purchase volume trends.
Conclusion: These insights suggest a strategic opportunity to target the 46-55 age group with more specific marketing strategies and to specifically focus campaigns on the male audience to maximize engagement and sales.
Key Analysis: Trends in Product Purchases Over Time reveal a significant spike in the 'Others' product category in May 2021, with purchases exceeding 50,000 units. This surge was likely influenced by major campaigns, special promotions, or seasonal factors. Additionally, 'Mobiles & Tablets' and 'Men's Fashion' categories showed peak performance in December 2020 and May 2021, respectively.
Tools Used:
- Python: Data analysis and preprocessing.
- Pandas: Data manipulation and aggregation.
- Matplotlib & Seaborn: Visualization of purchase trends.
Conclusion: To capitalize on these trends, maximize focus on leading categories like 'Others' and 'Men's Fashion' during their high-potential months. Conduct in-depth analysis of the May 2021 spike to identify drivers that can be replicated. Optimize distribution and stock during predicted peak periods and implement specific campaigns to maintain the performance of leading categories.
Key Analysis: The Global Sales Distribution chart visually represents sales volume across different geographical regions, with a color gradient highlighting higher (red) versus lower (blue) sales. The US Map provides a baseline for understanding sales distribution within the United States.
Tools Used:
- Python: Data analysis and preprocessing.
- Pandas: Data manipulation and aggregation.
- Matplotlib & Seaborn: Visualization of sales distribution.
- Geopandas: Geospatial data handling and plotting.
Conclusion: This visualization offers a comprehensive overview of sales performance across geographies, enabling stakeholders to pinpoint high-performing regions and identify potential new markets for growth. The detailed maps serve as a foundation for evaluating location-specific sales metrics, facilitating targeted decision-making for each state or region.
Key Analysis: The Monthly Quantity Ordered Trend shows notable fluctuations, with a sharp increase in December 2020 (102,520 units), followed by declines in January and February 2021. Significant growth was observed in March 2021 (74,889 units) and April 2021 (109,497 units), indicating potential seasonal effects or promotional successes.
Tools Used:
- Python: Data analysis and preprocessing.
- Pandas: Data manipulation and aggregation.
- Matplotlib & Seaborn: Visualization of sales trends.
Conclusion: Monitoring these monthly trends is crucial for optimizing inventory management and identifying key growth periods. This allows for proactive planning to capitalize on peak sales months and mitigate impacts during slower periods.
Key Analysis: Based on the Total Quantity Ordered by Region analysis, the Southern region recorded the highest quantity ordered (199,659 units), followed closely by the Midwest (171,968 units).
Tools Used:
- Python: Data analysis and preprocessing.
- Pandas: Data manipulation and aggregation.
- Matplotlib & Seaborn: Visualization of sales trends.
Conclusion: The Southern and Midwest regions are identified as primary drivers of product demand. Implementing targeted strategies in these high-performing areas can maximize growth opportunities and ensure efficient product distribution to meet regional demand.
Key Analysis: The Distribution of Product Variant Sales by Region chart illustrates that the South continues to record the highest sales volume for various product variants compared to other regions. Conversely, the West and Northeast regions exhibit lower overall sales volumes. Notably, categories like 'Superstore' and 'Books' maintain strong contributions across nearly all regions.
Tools Used:
- Python: Data analysis and preprocessing.
- Pandas: Data manipulation and aggregation.
- Matplotlib & Seaborn: Visualization of sales trends.
- Plotly: Interactive visualization.
Conclusion: While strengthening focus on the dominant Southern region, there are clear opportunities to enhance marketing and distribution strategies in the West and Northeast to boost sales volumes. Continuing to support high-contributing categories like 'Superstore' and 'Books' across all regions remains vital for overall sales performance.
This comprehensive project leverages a wide array of tools and technologies for in-depth data analysis, predictive modeling, and visualization:
- Programming Languages: Python
- Core Data Libraries: Pandas, NumPy
- Machine Learning & Statistical Models: Scikit-learn (Logistic Regression, Random Forest, KMeans Clustering), ARIMA (Time Series Analysis), Regression Analysis
- Data Visualization: Matplotlib, Seaborn, Plotly
- Geospatial Analysis: Geopandas
- Concepts: Churn Analysis, Demand Forecasting, Customer Segmentation, Price Elasticity, Time Series Analysis, Geospatial Analysis