Demand Planning
Combine statistical forecasting, external signals and accountable collaboration into a demand signal that supply, inventory and production teams can execute.
OpsFactor separates the level at which the model learns from the level at which people review, collaborate and execute.
01 · STATISTICAL & AI BASELINE
Run classic statistical models and AutoFit backtests at material-location, cluster or configured aggregate levels. Advanced paths incorporate external series and AI forecasting models, then disaggregate the result while preserving the individual demand profile of products, customers and channels.
02 · CLUSTERS & HIERARCHIES
Cluster products or customers by relevant demand behavior, not only by commercial taxonomy. Hierarchical and top-down paths make sparse lower-level series usable without forcing every item through the same model.
03 · CONFIGURABLE COLLABORATION
Create as many planning layers as the process needs: marketing events, engineering input for new items, sales-supervisor collaboration, management review and final consensus. Each role works through controlled Key Figures in Planning Books or governed bulk updates.
04 · FORECAST VALUE ADDED
Evaluate accuracy and bias for the statistical baseline and for each collaboration stage. Planners can see whether a contribution adds information or destroys value before it becomes the demand signal sent to execution.
Sell-in and sell-out histories can coexist by channel, while calendars, events, product introductions and succession rules keep the model aligned with the business context.
The numbered rows map the detailed capability set directly to each edition: start with a transparent statistical workflow and add advanced models, enterprise signals and governed collaboration as complexity grows.
| Capability | Community | Pro / Enterprise |
|---|---|---|
| Classic statistical forecasting and sell-out history | Included Standard models, historical split and planning Key Figures. | Included, with managed execution and support. |
| Cluster-level forecast and manual model configuration | Included Configured forecast scope, classic models and controlled review surfaces. | Included, with managed governance and scale. |
| Point 03AAdvanced statistical models, AutoFit and AI models | Not included Community uses the classic statistical catalog. | Enterprise Seasonal Naive, STL, Prophet, ETS, TBATS, automated model selection and AI forecasting models such as Chronos. |
| Point 03BExternal series and profile-aware disaggregation | Historical-sales split from the configured forecast level. | Enterprise External support series and covariates can inform the forecast; hierarchical and forecast-proportion paths preserve individual product, customer and channel profiles during disaggregation. |
| Sell-in, hierarchical reconciliation and event uplift | Not included | Enterprise Additional histories, reconciliation and preprocessing. |
| Point 07AConfigurable collaboration workflow by planning layer | Planning Books and standard Key Figures provide the basic collaboration surface. | Enterprise Governed stages can separate marketing events, engineering input, sales reviews, management approval and final consensus. |
| Point 07BAccuracy and Forecast Value Added by collaboration stage | Baseline and scenario comparison. | Enterprise Accuracy and bias can be evaluated by workflow stage to show which contributions add or destroy forecast value. |