This reference provides the detailed decision logic, optimization models, method selection trees, and segmentation methodologies for inventory demand planning a
The Decision Frameworks skill provides a structured approach to inventory demand planning for multi-location retailers managing hundreds of SKUs across categories like grocery, seasonal, and promotions. It guides practitioners through categorizing demand patterns using metrics such as coefficient of variation, average demand interval, and seasonal strength, then selecting appropriate forecasting methods and safety stock models based on these archetypes. The framework incorporates optimization protocols and model switching rules to maintain forecast accuracy over time, adapting to changes like new product launches or promotional impacts.
This skill is designed for supply chain analysts, demand planners, and inventory managers working in retail or wholesale environments with complex, multi-location SKU assortments. It’s especially relevant for practitioners responsible for forecasting and stocking decisions where demand varies by category type, seasonality, and product lifecycle stage. Growth leads or agency strategists supporting retail clients with inventory optimization challenges will find this framework valuable for aligning inventory levels with service goals while controlling holding costs.
First, practitioners calculate key demand metrics—such as coefficient of variation, average demand interval, and seasonal autocorrelation—using historical sales data to identify the product’s demand archetype (e.g., smooth, seasonal, intermittent). Next, they assign forecasting methods tailored to the archetype, like Holt-Winters for seasonal products or Croston’s method for intermittent demand, tuning parameters via grid search to minimize forecast error metrics like WMAPE. Then, they apply model switching rules, monitoring tracking signals and test periods to decide when to re-evaluate or switch forecasting methods. Finally, they use safety stock models adjusted for demand and lead time variability to set optimal inventory buffers aligned with target service levels.
How often should I reassess the demand archetype for a SKU? Quarterly reassessment is recommended to detect shifts such as from smooth to trending demand patterns. What triggers a model switch in forecasting? A WMAPE improvement greater than 10% on holdout data after an 8-week parallel test or significant tracking signal breaches prompt model review. How do I avoid overfitting when optimizing smoothing parameters? Prefer parameter sets with lower smoothing coefficients when WMAPE improvements exceed 5 percentage points on holdout versus training data, as excessive gains indicate overfitting.
Attach this skill to your Metaflow agent task when inventory demand planning requires nuanced decision logic and forecasting model selection based on detailed demand analytics. The agent will load the framework on demand to categorize demand patterns, assign forecasting methods, and recommend safety stock levels tailored to your SKUs’ characteristics. Expect automated evaluation of model performance with periodic reassessment and parameter tuning to maintain forecast accuracy over time—this ensures your inventory decisions balance service levels and carrying costs effectively.
For broader context, see our roundup of marketing skills claude, and read Claude Code workflows for marketing agencies for related setup guidance.