Demand Planning Essentials

Cheatsheet Content

Demand Planning Overview Definition: Process of forecasting customer demand to make informed supply chain decisions. Goal: Balance supply with demand, minimize costs, maximize customer satisfaction. Key Inputs: Historical sales, market intelligence, promotional plans, economic indicators. Demand Forecasting Methods 1. Qualitative Methods Description: Subjective, based on expert opinions, useful for new products or limited data. Techniques: Delphi Method: Structured communication technique for group consensus. Market Research: Surveys, interviews, focus groups. Sales Force Composite: Aggregating individual sales forecasts. Executive Opinion: Senior management's collective judgment. 2. Quantitative Methods Description: Objective, data-driven, based on historical data and statistical models. Time Series Models: Moving Average (MA): Simple average of past demand. $F_t = \frac{1}{N} \sum_{i=1}^{N} A_{t-i}$ Weighted Moving Average (WMA): Assigns different weights to past observations. $F_t = \sum_{i=1}^{N} w_i A_{t-i}$ Exponential Smoothing (ES): Weights recent observations more heavily. $F_t = \alpha A_{t-1} + (1-\alpha) F_{t-1}$ Seasonal Decomposition: Separates trend, seasonal, and random components. ARIMA (Autoregressive Integrated Moving Average): Handles trends, seasonality, and autocorrelation. Causal Models: Regression Analysis: Identifies relationships between demand and independent variables (e.g., price, promotions, economy). $Y = \beta_0 + \beta_1 X_1 + \epsilon$ Econometric Models: More complex regression, often with multiple equations. Accuracy Metrics Mean Absolute Deviation (MAD): Average of the absolute errors. $MAD = \frac{\sum |A_t - F_t|}{N}$ Mean Squared Error (MSE): Average of the squared errors. $MSE = \frac{\sum (A_t - F_t)^2}{N}$ Mean Absolute Percentage Error (MAPE): Average of absolute percentage errors. $MAPE = \frac{1}{N} \sum \left| \frac{A_t - F_t}{A_t} \right| \times 100\%$ Bias (Mean Forecast Error): Indicates consistent over/under-forecasting. $Bias = \frac{\sum (A_t - F_t)}{N}$ Tracking Signal: Monitors if forecast bias is within acceptable limits. $TS = \frac{Cumulative Error}{MAD}$ Forecasting Process Steps Data Collection: Gather historical sales, market data, internal plans. Data Cleansing: Identify and correct outliers, missing data, and anomalies (e.g., promotions, stock-outs). Method Selection: Choose appropriate qualitative/quantitative methods based on data availability, product lifecycle, and business context. Forecast Generation: Apply selected models to create baseline forecasts. Collaboration & Consensus: Integrate market intelligence from sales, marketing, finance, and operations. Review & Adjustment: Adjust forecasts based on consensus meetings and expert judgment. Performance Monitoring: Track actual demand against forecasts and analyze accuracy metrics. Continuous Improvement: Refine models and processes based on feedback and performance. Key Factors Influencing Demand Price: Elasticity of demand. Promotions/Discounts: Short-term spikes. Competitor Actions: Market share shifts. Economic Conditions: GDP, inflation, consumer confidence. Seasonality: Regular, predictable patterns (e.g., holidays, weather). Trends: Long-term upward or downward movements. Product Lifecycle: Introduction, growth, maturity, decline. External Events: Pandemics, natural disasters, supply chain disruptions. S&OP (Sales & Operations Planning) Definition: Integrated business management process to align demand and supply. Objective: Create a single, unified plan across all functions to balance demand, supply, and financial goals. Key Meetings: Product Review Demand Review Supply Review Pre-S&OP Meeting Executive S&OP Meeting Benefits: Improved forecast accuracy, better inventory management, enhanced customer service, increased profitability. Inventory Management Integration Safety Stock: Buffer inventory to protect against demand and supply variability. $Safety\ Stock = Z \times \sigma_L$, where $Z$ is service level factor, $\sigma_L$ is std dev of demand during lead time. Reorder Point (ROP): Level at which new order is placed. $ROP = (Average\ Daily\ Demand \times Lead\ Time) + Safety\ Stock$ Economic Order Quantity (EOQ): Optimal order quantity to minimize total inventory costs. $EOQ = \sqrt{\frac{2DS}{H}}$ (D=annual demand, S=order cost, H=holding cost) Challenges in Demand Planning Data Quality: Inaccurate or incomplete historical data. Volatility: Highly unpredictable demand patterns. Long Lead Times: Makes forecasting further out more difficult. New Product Introductions: Lack of historical data. Promotional Impact: Difficulty in quantifying effects. Human Bias: Over- or under-forecasting due to personal opinions.