Predicting how many items you will sell isn't one-size-fits-all. The exact same item can sell very differently depending on where it is located, and different products sell in completely different ways. To give you the clearest predictions, Inventory Capital Solutions (ICS) from FDC Solutions runs a Forecast Model Selection Tournament for every single combination of part number and location.
What is the Forecast Model Selection Tournament in ICS?
The Forecast Model Selection Tournament is an automated competition where every enabled forecast method generates its own prediction for an item using historical sales data. ICS measures how accurately each forecast method would have predicted past demand and calculates an error score. The formula with the lowest error wins the tournament for that specific part and location.
By running this selection at a granular level, ICS pinpoints stock-out risks before they happen and provides purchasing recommendations tailored specifically to every individual part number at every location.
Breakdown of ICS Forecast Methods
The table below contains a description of the forecast methods used in ICS.
Forecast Method | Primary Application | Key Advantage |
Weighted Average | Stable items or new parts with short sales histories (<6 months). | Serves as a safe fallback; gives more weight to recent sales while smoothing noise. |
Moving Average | Products with flat, steady, and unchanging demand. | Simple and reliable; keeps order recommendations steady across set time windows. |
Average | Mature products with sales spread evenly across all 12 months. | Calculates a pure, unweighted baseline when demand shows zero volatility. |
Single Exponential Smoothing | Items with stable demand where recent trends need quick inclusion. | Blends past forecasts with recent actuals to quickly adapt to minor demand changes. |
Double Exponential Smoothing | Items experiencing gradual upward or downward trends over time. | Adaptive level and trend features self-correct for steady growth or decline. |
Seasonal Demand | Highly seasonal items or slow-moving seasonal inventory. | Uses percentage factors to match yearly spikes while preserving timing patterns. |
Sporadic and Intermittent Demand | Slow movers and spare parts with fewer than 3 sales hits per year. | Prevents zero-sale months from distorting future ordering signals. |
Why Automated Forecast Selection Adds Value
Manual forecast selection or static "one-size-fits-all" rules inevitably fail because product demand patterns constantly evolve. Automating the selection of the forecast method adds immediate value to your inventory planning.
Eliminates Human Bias and Manual Work. Planners no longer need to manually review spreadsheets or guess which math model to apply to thousands of part numbers.
Adapts to Product Lifecycles. As a part transitions from a fast-growing new release to a stable mover, or eventually to a sporadic spare part, the tournament automatically switches to the best-fitting algorithm.
Optimizes Capital and Service Levels. Picking the most accurate algorithm at the location level prevents costly overstocking while protecting against stock-outs, keeping capital productive and customers satisfied.
Ready to optimize your inventory planning process?
The Forecast Model Selection Tournament embedded within Inventory Capital Solutions (ICS) from FDC Solutions continuously tests and selects the most accurate forecast method for every combination of part number and location, protecting your cash flow and boosting service levels.
Complete the form below to discover how ICS addresses your inventory management challenges to transform your inventory into a competitive advantage.