Inventory Forecasting Model

The Goal: Forecasting Next Month's Inventory

This project was for a children's clothing company. The main goal was to create an inventory forecast for **January 2025** by using the sales history from all of 2024. I used a fictional dataset for this public version since the real data is confidential.

My objectives for this task were:

• To analyze the 2024 data to find out which dresses were the fast-moving, moderate-moving, and slow-moving items.
• To use that analysis to build a weighted average model that could accurately forecast the inventory needed for January 2025.

My Process: How I Built the Model

I built the entire forecasting model in Excel. Here are the steps I took to get the final forecast:

• First, I used a pivot table to filter the data down to exactly what I needed: only dresses from the "Mini Miss" department sold in the webstore at a discounted price.
• Then, I analyzed the sales history for 2024 to find the average sales for different seasons, which I called lenient, moderate, and peak months.
• I computed a weighted average for each dress. I gave more weight to the lenient months (60%) and less to the peak months (10%) because I wanted a more conservative and accurate forecast for January, which is a lenient month.
• Next, I classified each dress as fast, moderate, or slow-moving based on its weighted average score.
• Finally, I applied a simple multiplier to the weighted average. Fast-movers got a small boost (x1.2), while slow-movers were reduced (x0.8), to get the final forecast number for each dress.