The Forecasting tab in Recurrency’s Demand Planning module offers a comprehensive view of your stockable items, their historical usage, and future demand projections through Machine-Learning Driven Forecasting. By analyzing past usage patterns, Recurrency generates precise forecasts to help you manage inventory more efficiently.
💡 It's important to note that while the Forecasting tab provides key insights into item behavior, you're not expected to take direct action here. Instead, forecasting serves as a building block for Recurrency’s Planning Recommendations, which guide your replenishment decisions. The forecasts give you a foundation of trust in Recurrency’s predictions, so you can feel confident when reviewing and accepting planning recommendations later on.
How Forecasting Works
Recurrency leverages machine learning to analyze historical usage data, identifying demand patterns and generating forecasts. This is particularly valuable for understanding seasonal trends, spikes due to holidays, and irregular demand. Data is refreshed weekly, ensuring you have the most up-to-date information to work with.
For ERPs like P21, which track usage directly, Recurrency pulls the data seamlessly. If your ERP doesn't manage usage or if you want increased flexibility in usage calculations, Recurrency computes it based on your inputs, providing flexibility for teams that need more control over how usage is calculated. Each month, Recurrency imports or calculates the prior month’s usage data, ensuring your forecasts are as accurate as possible.
💡 Note: Non-stockable items are not included in Recurrency forecasts as we are not proactively planning for them.
Smart Forecasting Explained
Recurrency’s Smart Forecasting uses advanced machine learning algorithms that consider factors such as:
Seasonality: Monthly and yearly trends
Holiday Impacts: Adjusts for U.S. holiday effects
Intermittent Demand: Handles inconsistent demand patterns
Outlier Filtering: Ignores outliers flagged in the ERP or in historical data
With Smart Forecasting, you can easily adjust the time horizon (e.g., 3 or 6 months), filter by supplier, product group, location, or demand pattern, and download data for deeper analysis. The Forecasting tab is designed to help you build trust in how Recurrency predicts item behavior, offering clear signals about where you may need to spend more time reviewing or adjusting.
Our forecasting models use up to 10 years of historical data. While we prioritize recent data, the historical provides trends for the item and is taken into account when generating the forecasts.
Demand Patterns and Predictability
Recurrency assigns a Demand Pattern to each item, analyzing how usage fluctuates over time. This helps you understand whether demand is smooth, erratic, or sporadic. Demand patterns are a crucial part of forecasting and inventory management, enabling you to align your stock levels with actual demand behavior. Learn more about Demand Patterns here.
Additionally, Predictability Scores help you gauge the reliability of these forecasts. Items are categorized as Very High, High, Medium, or Low based on how consistent their past demand has been. Higher predictability means the item is easier to forecast, allowing you to trust the replenishment recommendations with confidence. Lower predictability suggests more irregularity, which may require closer review. Learn more about Predictability here.
Visualizing Trends & Key Data
To dive deeper into item behavior, click on any forecast graph to view detailed trends, such as monthly usage history, sales, orders, purchase orders (POs), availability, and costs. You can zoom in and out to see more granular or long-term views, allowing you to spot trends like large usage shifts and adapt your forecasts accordingly.
Everything from the Forecasting tab is also available in the context panels of other tabs within the Demand Planning module, ensuring you have access to all the critical insights when making decisions.
Common Use Cases
Review Low-Predictability Items
Use the Predictability filter to identify items that need more attention. For example:Filter Predictability = Low
Supplier = A high priority supplier
Review forecasts for these items to validate assumptions and adjust stock levels accordingly.
Explore Item Trends Over Time
Zoom in on individual item-location pairs to understand historical demand patterns and how Recurrency adapts forecasts based on real-time usage updates.