Recurrency uses Demand Patterns to group items based on consistency of orders and variance in demand quantity based, and uses these to forecast in a more accurate manner than other solutions. These forecasts also inform Recurrency’s dynamic min/max recommendations for each item.

Although our forecasting model utilizes up to 10 years of historical data to identify trends and generate forecasts, our demand pattern classification relies solely on the past 12 months of usage data.

Demand Patterns are determined in Recurrency, and can be broken down into items with low demand, and items with high demand. In the platform, you can hover over the icon next to each demand pattern to see its description.

Items without much history:

Items with sufficient history:

Recurrency uses historical usage data to generate precise forecasts for each item-location pair, helping you stay ahead of demand fluctuations. We cover all demand patterns, from high-demand items like smooth, erratic, and sporadic, to low-demand items like new and no-demand items. However, sparse items—those with only six months of usage in the past year—are treated differently.

Since sparse items are often overstocked or lead to stockouts when planned manually, Recurrency approaches them cautiously. Instead of displaying a traditional forecast, we show the average monthly demand. While our platform still calculates the forecast in the background, this method gives you insight into how much will likely be reordered without predicting when it will happen, ensuring more informed decisions for these unpredictable items.

You can learn more about Sparse Items in Recurrency here.