Skip to content

README.md

Latest commit

 

History

History
120 lines (83 loc) · 4.6 KB

README.md

File metadata and controls

120 lines (83 loc) · 4.6 KB

📦 NZ Retail Inventory Optimization & Forecasting Dashboard

🧩 Overview

This project addresses a critical challenge in New Zealand’s retail sector: working capital strain caused by excess inventory due to misaligned supply and demand. By leveraging forecasting models, integrating multiple datasets, and developing a professional Power BI dashboard, this solution empowers decision-makers to optimize inventory and uncover potential cost savings.

🔍 Problem Statement

New Zealand retailers often hold surplus inventory, tying up significant working capital. This issue is driven by inaccurate demand predictions and delayed supply alignment. The need was clear: a data-driven approach to identify inventory inefficiencies and enable proactive decision-making.

✅ Objectives

  • Forecast future sales using historical data.
  • Identify gaps between forecasted demand and actual inventory.
  • Provide actionable insights for supply chain and finance teams.

⚙️ Solution Approach

🧠 Forecasting & Data Modeling

  • Developed a time series forecasting model using Facebook Prophet in Python.

  • Forecasts were generated at a regional and industry level (Retail & Accommodation sectors).

  • Integrated outputs with historical retail sales, inventory levels, cost data, and capacity constraints.

  • Retail Chain Analysis

A full retail analytics project focused on sales forecasting, regional trends, and supply chain optimization using Python and real New Zealand retail data.

[Open In Google Colab]

📊 Python-Based Analytical Pipeline

  • Built a forecasting-driven inventory optimization model using Python, combining time series predictions with business metrics.

  • Prophet Forecasting: Used Facebook Prophet to forecast inventory demand at a quarterly level for all regions.

  • Forecast Accuracy Metrics: Computed MAE, RMSE, and MAPE to evaluate model performance.

  • Sales-Proportional Inventory Allocation: Allocated forecasted inventory by region based on regional sales share.

  • Inventory Efficiency Metrics: Calculated Excess Inventory, Holding Costs, and Forecast Error at region level.

  • Top 10 Region Matrix: Created a clean KPI matrix highlighting regions with highest excess stock.

  • Visualizations with Plotly: Generated interactive heatmaps, bar charts, and KPI tables to visualize inventory inefficiencies.

🛠 Tech Stack

Tool Purpose
Python (Prophet) Sales forecasting (Time Series)
Pandas & NumPy Data cleaning and transformation
Stats NZ Datasets Official historical data sources

📈 Results & Impact

  • Delivered a data-rich, decision-ready analytics solution that:

  • Uncovered critical inefficiencies in inventory allocation, highlighting regions and quarters with excess stock and missed demand signals.

  • Quantified cost-saving potential through measurable reductions in holding costs and excess inventory, enabling proactive budget realignment.

  • Empowered cross-functional teams with clear, visual diagnostics—supporting agile responses to forecast mismatches and regional disparities.

  • Strengthened data-driven strategic planning by linking supply chain decisions to working capital optimization, financial health, and operational resilience.

📂 Project Structure

NZ-Retail-Inventory-Optimization/
│
├── data/
│   ├── Annual_Enterprise_Survey_2023.csv
│   ├── CPI.csv
│   ├── Forecast.csv
│   ├── Forecast_Model.csv
│   ├── Inventory.csv
│   ├── Regional_Sales.csv
│   ├── Retail_Sales.csv
│   ├── Retail_Trade_Sales_and_stocks_by_industry_1995–2025.csv
│   ├── Wholesale.csv
│   └── nz_retail_inventory_optimized.csv
│
├── forecasting/
│   └── supply_chain_forecasting.ipynb
│
├── powerbi/
│   ├── NZ_Inventory_Optimization.pbix
│   └── pbix_visual_map.md          
│
├── visuals/
│   ├── dashboard_screenshots/
│   │   ├── regional_heatmap.png
│   │   ├── top10_excess_bar.png
│   │   └── executive_summary_table.png
│   └── final_presentation_figures/
│
├── documentation/
│   ├── README.md
│   ├── insights_summary.md
│   └── markdown_tables.md          
│
├── reports/
│   └── NZ_Retail_Case_Report.pdf
│
└── requirements.txt