Operations Management and Forecasting Essentials

Posted on Mar 9, 2026 in Business Administration and Management (BAM)

1. Forecasting Definitions

Forecast: Estimate of future demand using past data and judgment.
Independent demand: External demand that must be forecasted.
Dependent demand: Derived from BOM; handled by MRP.

Time Series: Data over time with components: Trend (T), Seasonality (S), Cyclical (C), and Random (R).
Stationary series: Mean and variance are constant; no trend or seasonality.
Error (eₜ): Aₜ − Fₜ.

Forecast characteristics: Unbiased, low error, and stable.
Horizons: Short term (days/weeks) vs. Long term (years).
Methods: Judgmental (Delphi, Sales force composite, Exec opinions) vs. Quantitative (Time series, Causal/Regression).

2. Forecasting Formulas

  • Naïve: Fₜ₊₁ = Aₜ
  • Simple Moving Avg (n): Fₜ₊₁ = (Aₜ + Aₜ₋₁ + … + Aₜ₋ₙ₊₁)/n
  • Weighted MA: Fₜ₊₁ = Σ wᵢAᵢ (where Σwᵢ = 1)
  • Exponential Smoothing: Fₜ₊₁ = αAₜ + (1−α)Fₜ (α ∈ (0,1); higher α = more responsive)
  • MAD: Σ|eₜ| / n
  • MAPE: (100/n) Σ(|eₜ|/Aₜ)
  • MSE: Σeₜ² / n
  • Regression: Y = a + bX (b = Cov(X,Y)/Var(X); a = Ȳ − bX̄)
  • R²: Explained variation / Total variation
  • Seasonal Index: Actual / Trend
  • Forecast with seasonality: Trend × Seasonal Index

3. Forecasting Frameworks

Seven Steps in Forecasting

  1. Define purpose
  2. Select items
  3. Choose horizon
  4. Select model
  5. Gather data
  6. Make forecast
  7. Validate

Decomposition Steps

  1. Estimate trend (regression)
  2. Compute seasonal indices
  3. Deseasonalize
  4. Forecast components
  5. Reseasonalize

4. Common Forecasting Mistakes

  • Using MAPE when Aₜ≈0 (distortion)
  • Comparing MAD across different scales
  • Forgetting Σwᵢ=1 in WMA
  • Choosing α without testing
  • Ignoring residual patterns
  • Using regression without checking R²
  • Assuming correlation implies causation
  • Not deseasonalizing before regression
  • Lag problem in moving average
  • Overfitting long-term data

5. Operations Management Definitions

Operations Management (OM): Design, operation, and improvement of transformation systems.
Transformation Process: Inputs (M, L, C, Info, Energy) → Process → Outputs (Goods/Services) + Feedback.
Supply Chain Management (SCM): Integration of Suppliers, Manufacturers, Distributors, Retailers, and Customers.
Value Proposition (VP): Why customers buy (Cost, Quality, Speed, Flexibility, Innovation).
Productivity: Output / Input.
Order Qualifier (OQ): Minimum requirement to compete.
Order Winner (OW): Differentiator that wins the order.

6. Operations Core Formulas

  • Productivity: Output / Input
  • Labor Productivity: Output / Labor hours
  • Multi-factor Productivity: Output / (Labor + Capital + Material + Energy)
  • Capacity Utilization: Actual Output / Design Capacity
  • Efficiency: Actual Output / Effective Capacity
  • Break-even Q: FC / (P − VC)

7. Quality Management Definitions

Quality: Customer perception; “Fitness for use” (Juran).
VOC: Voice of Customer (needs/expectations).
CTQ: Critical to Quality (measurable attributes).
TQM: Org-wide quality management.
Six Sigma: Philosophy and methods to reduce variation and eliminate defects.
Cost of Quality: Price of Conformance (Prevention + Appraisal) + Price of Non-Conformance (Internal + External failure).

8. Quality Formulas

DPMO: (Defects / (Units × Opportunities per unit)) × 10⁶
Control Limits: UCL = μ + 3σ; LCL = μ − 3σ