Comprehensive retail data analysis for an Indian ethnic wear brand — turning raw sales, P&L, inventory, and logistics data into actionable business strategy.
| Dimension | Coverage |
|---|---|
| Business Domain | Indian Ethnic Wear Retail (D2C + B2B + International) |
| Period | Q3 2022 Planning |
| Data Sources | Amazon Sales, P&L, Inventory, International Sales, Logistics |
| Core Themes | Profitability, Demand, Inventory Health, Logistics Cost, Seasonality |
| Output | 7 analytical sections + Insights + Actionable Recommendations |
Modern Heritage is a multi-channel Indian ethnic wear brand selling across Amazon India, international marketplaces, and B2B channels. Despite broad market reach, the company lacked an integrated view of its profitability, inventory efficiency, and logistics costs.
This project was commissioned to bridge that gap — integrating four disparate datasets (sales, profit & loss, inventory, and logistics) into a unified analytical framework that answers the most pressing business questions ahead of Q3 2022 planning.
The analysis moves beyond surface-level revenue tracking to uncover where real margin is made, where demand is concentrated, and where operational costs can be reduced.
Modern Heritage operates at the intersection of fashion, e-commerce, and traditional retail — a space where product mix decisions, logistics partnerships, and inventory allocation can make or break profitability.
The client's core challenges were:
- No unified view of profitability across product categories after accounting for transfer prices
- Unclear demand geography — which states drive volume, and which are underserved
- Inventory risk — dead capital tied up in slow-moving SKUs alongside stockouts of high-velocity items
- Unvalidated international expansion — whether cross-border logistics costs were justified by revenue
- Logistics partner uncertainty — no data-driven comparison between Shiprocket and INCREFF
This analysis was designed to answer seven core business questions:
- Which product categories are the most profitable — not just the most sold?
- Which states generate the highest order volume and should be prioritized for inventory placement?
- What is the inventory health across SKUs, and where is capital at risk?
- Is international revenue worth the logistics overhead compared to domestic sales?
- Which logistics partner — Shiprocket or INCREFF — offers better unit economics at current volumes?
- What does the distribution status breakdown look like, and what operational decisions does it suggest?
- Are there seasonal or monthly sales patterns that should drive inventory and marketing cycles?
| File | Description |
|---|---|
Amazon Sale Report.csv |
Order-level Amazon India sales data including SKU, category, quantity, amount, state, and order status |
P&L Report |
Transfer price (TP1) data per SKU enabling true margin calculation |
Inventory Report |
Stock levels per SKU |
International Sales Report |
Gross revenue from international channels |
Logistics Comparison |
Per-unit outbound cost rates for Shiprocket vs. INCREFF |
| Tool | Purpose |
|---|---|
| Python 3 | Core analysis language |
| Pandas | Data loading, cleaning, merging, aggregation |
| Plotly Express | Interactive bar charts, line charts, pie charts |
| Seaborn | Statistical scatter plots, styled grid visuals |
| Matplotlib | Supporting visualizations |
| Jupyter Notebook | Reproducible analysis environment |
Raw data from multiple sources required standardization before analysis could begin.
Key cleaning steps performed:
- Redundant column removal:
Unnamed: 22in the Amazon Sales Report was entirely null and dropped. - Missing value handling: The
Amountcolumn had 6.04% null values, attributed primarily to cancelled orders, and handled accordingly. - Date parsing: The
Datecolumn was converted to properdatetimeformat usingpd.to_datetime()with error coercion. - SKU normalization: SKU formatting was inconsistent across files; standardized to enable reliable joins across the sales, P&L, and inventory datasets.
- P&L column standardization: Column headers were stripped and uppercased to ensure consistent merging.
- International revenue cleaning: Gross amount values contained comma-formatted strings and were converted to numeric with null-fill for missing entries.
- Master dataset construction: A merged
df_masterwas built from the Amazon sales and P&L data to enable profit calculations at the order level.
Revenue was first analyzed at the category level to identify top-selling items. Sets and Kurtas led in gross revenue. However, after deducting the transfer price (TP1) per unit, the Set category significantly outperforms all others in net profit, making it the highest-priority product line for the business.
Categories like Saree, Dupatta, and Bottom showed comparatively weaker margin contribution, suggesting a need to deprioritize capital allocation toward them.
Order volume was mapped across ship-to states to identify geographic demand concentration. Maharashtra leads as the top consumer state, followed by Karnataka and Tamil Nadu. Together, these three states represent the highest concentration of demand.
Gujarat, Andhra Pradesh, West Bengal, Kerala, and Delhi were identified as passive consumers — markets with latent potential or lower brand penetration.
A scatter analysis was conducted correlating stock levels with sales velocity (units sold) per SKU. The results showed a moderately positive correlation — products with higher stock availability tended to achieve higher total sales, confirming that inventory availability supports revenue.
However, the analysis also flagged a segment of SKUs with high stock but low sales velocity, indicating potential dead capital. Simultaneously, high-velocity SKUs risk stockouts if not proactively restocked. The analysis identified the top 50 at-risk SKUs on both ends of the spectrum.
Revenue was aggregated separately for domestic (Amazon India) and international channels. The analysis revealed that 82.7% of total revenue is generated domestically, with the international market contributing only 17.3%.
Given the significantly higher logistics and operational overhead associated with international fulfillment, this comparison raises a clear strategic question: is the current investment in international infrastructure proportionate to the revenue it generates?
Per-unit outbound costs were compared across the two logistics partners:
| Partner | Outbound Cost per Unit |
|---|---|
| Shiprocket | ₹7.00 |
| INCREFF | ₹11.00 |
Applying these rates to total order volume demonstrated a material cost difference in favor of Shiprocket. While INCREFF offers more robust inventory management capabilities, Shiprocket is the more cost-effective choice at current operational volumes, unless INCREFF can offer bulk pricing comparable to Shiprocket's rates.
Order statuses were aggregated and visualized as a proportion chart. The analysis confirmed that a significant majority of orders are shipped successfully, while a notable proportion are cancelled. Smaller categories (below 2% share) were grouped into an "Other" bucket for cleaner visualization.
This breakdown informs operational decisions around cancellation rates, return logistics, and fulfillment reliability.
Monthly revenue was aggregated and plotted as a time-series to detect seasonality. The data clearly showed that revenue peaks during the festive season, a pattern well-established in Indian ethnic wear retail. This pattern has direct implications for inventory pre-stocking, marketing spend timing, and logistics capacity planning.
- Sets drive the highest margin — despite both Sets and Kurtas leading in raw revenue, the Set category dominates after transfer price deduction, making it the highest-ROI product line.
- Maharashtra, Karnataka, and Tamil Nadu are the primary demand hubs and should be the focus of inventory placement and targeted marketing.
- Dead capital risk exists in the inventory — several SKUs carry high stock with low sales velocity, tying up working capital inefficiently.
- 82.7% of revenue is domestic — international logistics costs may not be justified at current revenue contribution levels.
- Shiprocket is the more profitable logistics partner at current order volumes, offering a ₹4/unit cost advantage over INCREFF.
- Sales are seasonal — the festive season drives revenue peaks, requiring advanced inventory build-up and marketing preparation.
- The P&L integration adds critical analytical depth — this analysis reveals true margin rather than surface-level sales figures, enabling better capital allocation decisions.
⚠️ Data Limitation Note: The cost data used in this analysis is from 2021. Inflationary trends in textile raw materials post-2021 may affect current margin calculations.
-
Inventory Realignment — Restock the top 50 high-velocity SKUs identified in the inventory-sales risk analysis to prevent stockouts and lost revenue.
-
Double Down on Sets — Reallocate capital and production focus toward the Set category, which shows the highest margin contribution. Reduce investment in underperforming categories such as Saree, Dupatta, and Bottom.
-
Logistics Partner Decision — Proceed with Shiprocket for domestic distribution to maintain lower per-unit outbound costs. Revisit INCREFF only if they can offer bulk rates competitive with Shiprocket's ₹7/unit pricing.
-
Geographic Marketing Focus — Prioritize marketing spend and promotional activity in Maharashtra, Karnataka, and Tamil Nadu. Consider targeted re-engagement campaigns in passive states like Gujarat and Delhi to expand the consumer base.
-
Festive Season Preparation — Use the monthly sales trend to time inventory builds and marketing campaigns ahead of seasonal peaks, reducing fulfillment delays and maximizing peak-season revenue capture.
-
International Market Review — With domestic revenue at 82.7% of total, conduct a cost-to-revenue audit of international operations before scaling them further.
Data Loading
↓
Initial Inspection (shape, dtypes, nulls)
↓
Data Preprocessing (cleaning, normalization, merging)
↓
Exploratory Data Analysis
├── Category Profitability
├── State-Wise Demand
├── Inventory Health & Sales Velocity
├── Domestic vs. International Revenue
├── Logistics Cost Comparison
├── Distribution Status Breakdown
└── Monthly Sales Trends
↓
Key Insights
↓
Business Recommendations
Prerequisites:
pip install pandas matplotlib seaborn plotly jupyterSteps:
# 1. Clone the repository
git clone https://github.com/<your-username>/modern-heritage-analysis.git
cd modern-heritage-analysis
# 2. Place the datasets in the /dataset folder
# - Amazon Sale Report.csv
# - P&L data
# - Inventory Report
# - International Sales Report
# 3. Launch the notebook
jupyter notebook Modern_Heritage_Growth_Analysis.ipynbThe notebook is self-contained and runs sequentially from data loading through to recommendations.
- Interactive Dashboard — Migrate key visualizations into a Streamlit or Power BI dashboard for stakeholder-facing reporting.
- SKU-Level Forecasting — Apply time-series models (Prophet or ARIMA) to forecast demand at SKU level for festive season planning.
- Customer Segmentation — Incorporate buyer-level data for RFM (Recency, Frequency, Monetary) segmentation to identify high-value customer clusters.
- Margin Refresh — Update transfer price data to 2023/2024 values to account for post-pandemic textile inflation and recalculate category margins.
- Return Rate Analysis — Deep-dive into cancelled and returned orders by category and state to quantify and reduce fulfillment losses.
Md Umair Alam
Data Analyst | Business Intelligence | Retail Analytics
📧 mdumairalam10@gmai.com
🔗 LinkedIn Profile
🐙 GitHub
This project was developed as part of a freelance data science engagement for Modern Heritage, covering Q3 2022 strategic planning.