Logistics Planning and Management: Horizons, Forecasts, Inventory
Planning horizons
Planning runs in three layers. Strategic (2–5+ years): network design - number, location, and mission of DCs, flow paths, outsourcing to 3PLs. Tactical (monthly to annual): S&OP, carrier contracts, inventory policy, seasonal capacity. Operational (daily/weekly): DRP, load plans, appointment scheduling. Questions often give a decision and ask the horizon - choosing a DC site is strategic; adding a shift is tactical.
Forecasting and S&OP
Forecasts are always wrong; the goal is measurable, planned-for error. Track bias (consistent over- or under-forecast) separately from random error (MAD, MAPE). Aggregate forecasts are more accurate than item-level ones - a trap is judging the process only at SKU level. S&OP is a monthly cross-functional cycle producing one demand plan and one supply plan, ending in an executive decision.
Inventory planning
Cycle stock covers demand between replenishments; EOQ trades order cost against carrying cost. Safety stock buffers variability: SS = z × σ of demand over lead time. DRP explodes DC requirements backward to plants and suppliers; MRP does the equivalent for dependent-demand components. Postponement (delayed differentiation) pushes customization downstream to cut finished-goods SKUs.
Worked example
Weekly demand standard deviation 40 units, lead time 4 weeks, 95% service (z = 1.65). Demand variability over lead time = 40 × √4 = 80. Safety stock = 1.65 × 80 = 132 units. The square root of lead time is the step candidates miss; multiplying by 4 gives a badly overstated answer.
Traps: EOQ ignores quantity discounts and freight breaks; safety stock is not cycle stock; zero average forecast error can still mean a useless forecast.