Demand Forecast
Top 10 SKUs by historical revenue — 4-week forward demand forecast with inventory risk flags.
Top 10 SKUs — Inventory Risk Table
Sorted by Days of Supply ascending (highest risk first)
| SKU | Description | Avg Wk Demand | 4W Forecast | Modelled Stock | Days of Supply | Risk |
|---|---|---|---|---|---|---|
| 23084 | Rabbit Night Light | 3,007 | 12,028 | 4,937 | 11.5 | ⚠ Reorder Risk |
| 79321 | Chilli Lights | 450 | 1,801 | 757 | 11.8 | ⚠ Reorder Risk |
| 84879 | Assorted Colour Bird Ornament | 966 | 3,865 | 2,661 | 19.3 | Monitor |
| 85123A | White Hanging Heart T-Light Holder | 884 | 3,537 | 2,775 | 22.0 | Monitor |
| 85099B | Jumbo Bag Red Retrospot | 912 | 3,647 | 3,478 | 26.7 | Monitor |
| 22423 | Regency Cakestand 3 Tier | 242 | 969 | 934 | 27.0 | Monitor |
| 23843 | Paper Craft, Little Birdie | 80,995 | 323,980 | 323,980 | 28.0 | ✓ Adequate |
| 47566 | Party Bunting | 186 | 745 | 1,153 | 43.3 | ✓ Adequate |
| 23166 | Medium Ceramic Top Storage Jar | 180 | 720 | 9,740 | 378.8 | ✓ Adequate |
| 22502 | Picnic Basket Wicker Small | 2 | 10 | 175 | 489.2 | ✓ Adequate |
Stock model: Modelled at 4× full-period average weekly demand (illustrative assumption — not observed inventory data). In a live deployment, this module would consume real-time inventory data from an ERP or WMS system (SAP, Oracle, Vinculum, or Increff for Indian retailers).
Threshold:<14 days = Reorder Risk (wholesale 2-week lead time). For Indian quick commerce dark stores (Blinkit, Zepto, Swiggy Instamart), this threshold would be ≤3 days.
Forecasting method: 4-week simple moving average (anchor date: 09 Dec 2011). A production model would use Holt-Winters exponential smoothing or Prophet to capture trend and seasonality.
Indian Retail Context
Quick Commerce (Blinkit, Zepto, Swiggy Instamart): Dark stores operate on 3–7 days of stock holding. A DoS threshold of ≤3 days would trigger immediate replenishment in this context. Stockout at a dark store results in direct revenue loss and order cancellation.
Kirana Supply Chain:India's informal distributor network is historically prone to stockouts driven by demand forecasting failures at the stockist level. FMCG majors (HUL, Nestlé India, Britannia) have invested in demand sensing tools to address this — the logic demonstrated here is a simplified analogue of those systems.
WMS Integration: Retailers using Increff or Vinculum for warehouse management already hold the SKU-level demand and stock data required to run this module in production. The forecasting logic would sit upstream of the replenishment planning layer, feeding reorder signals into the WMS.
Limitations
| Assumption | Detail | Impact if Wrong |
|---|---|---|
| Stock = 4× average weekly demand | Illustrative only — not observed inventory | If actual stock is lower, more SKUs would be at risk |
| 4-week SMA for forecasting | Ignores trend and seasonality | Q4 seasonal spike causes under-forecasting if run in November |
| No lead time variability | Assumes constant 14-day lead time | Variable lead times require safety stock buffer modelling |
| Top 10 by revenue, not margin | Revenue ≠ profit contribution | High-volume, low-margin SKUs may appear in the top 10 |