inventory-demand-planning

Match the method to the demand pattern: weighted moving averages for stable items, Holt's double exponential smoothing for trending items, Holt-Winters for seasonal items, and Croston's method for intermittent demand with many zero-demand periods. The Skill provides a decision table mapping each pattern to a primary method, fallback, and review trigger.

Use the combined formula SS = Z × √(LT_avg × σ_d² + d_avg² × σ_LT²) to capture both demand and lead-time variability. Vendors with lead-time coefficient of variation above 0.3 can require safety stock 40–60% higher than demand-only formulas suggest.

ABC ranks SKUs by margin contribution (A = top 20% driving 80% of margin), while XYZ ranks demand predictability by coefficient of variation (X < 0.5, Z > 1.0). The combined matrix drives policy: AX items get automated replenishment, while AZ items need human review each cycle.

Use analogous item profiling: select 3–5 similar items matched on price point, category, brand tier, and demographic, then borrow their demand variability and lifecycle curve. Commit a conservative initial buy around 60% of the analog-based forecast and add a 20–30% safety stock buffer for the first 8 weeks.

Promotions create a post-promo dip as customers consume stockpiled inventory, typically 30–50% of the incremental lift concentrated in the first 1–3 weeks. Strip promotional volume from baseline history and model the dip separately, defaulting to 40% of lift spread 60/30/10 across three weeks.

Flag a SKU when weeks of supply exceed 26, recent 13-week velocity is under 50% of its launch velocity, no promotions are planned, and no contractual commitment exists. Start with a 30% markdown for 4 weeks, escalate to 50% or liquidation, and set a hard exit date 8 weeks out.