Case Study: Building a Forecasting Tool to Reduce Inventory and Improve Cash Flow
Client Snapshot
A mid-sized home fashions company carried more than 30,000 SKUs across its product portfolio. The business was family-owned and operated with limited forecasting resources and a long-term contract with a commercial demand planning system.
The Challenge
The company held approximately eight million dollars in excess inventory. Inventory turns had fallen below two per year, limiting cash flow and tying up capital. Leadership believed they understood customer needs but were unsure why inventory was not selling through. Analysis showed the root issue was overbuying.
The forecasting software generated orders based solely on recent sales, without accounting for trends or seasonality. The system was complex, and the forecasting lead was unable to configure it effectively. With 30,000 SKUs to manage, a simpler tool was needed.
The Approach
A custom forecasting tool was developed in Microsoft Excel to run alongside the existing platform. Named FABIO (Forecasting Assistant with Blended Inventory Outlooks), the tool combined three standalone forecasting models and blended their outputs in four additional ways, creating a total of seven forecast options. FABIO used three years of historical data to forecast the most recent six-month period, then evaluated which model was most accurate based on actual results. That model was then used to forecast forward.
Users could adjust volumes manually for known events like promotions or closeouts. The tool also flagged high-volume or high-risk SKUs to prioritize review. Forecast output was formatted to upload via FTP into the company’s existing demand planning system, allowing continued use of downstream purchasing automation.
The Outcome
The Excel-based tool delivered rapid results. Forecast accuracy improved significantly. Inventory began moving, and new purchases better aligned with true demand. Cash flow improved, and the business avoided the need for new software.
Lesson Learned
The solution isn’t always expensive or off-the-shelf. A well-structured Excel tool, built with the right logic and clarity, can outperform larger systems when the problem is well understood.