Semi-Markov modelling

Semi-Markov modelling is a stochastic process where the transition probability between states depends only on the current state, but the sojourn time (the time spent in a state) can follow any distribution, not just the exponential distribution used in standard Markov models. This makes it superior for modeling business recovery processes where recovery times are rarely constant or memoryless. In the context of ISO 22301:2019, it provides a rigorous mathematical framework to validate Recovery Time Objectives (RTOs) by accounting for the actual variability observed in historical disruption data, ensuring that BCPs are based on realistic recovery trajectories rather than optimistic assumptions.

Implementation typically follows three steps: 1) State Definition: Categorize business functions into operational states (e.g., Normal, Degraded, Interrupted, Recovering) based on BIA findings. 2) Parameter Estimation: Use historical disruption data or industry benchmarks to fit sojourn time distributions (e.g., Lognormal or Weibull) and transition probabilities. 3) Scenario Simulation: Run Monte Carlo simulations to generate a distribution of recovery times, identifying the probability of meeting RTOs under different scenarios. For example, a Taiwanese semiconductor firm used this model to simulate a 20% reduction in workforce during a pandemic, optimizing their contingency staffing plan and improving RTO compliance by 25% within one year.

Three primary challenges exist: Data Scarcity (most SMEs lack structured disruption logs), Technical Complexity (the model requires statistical expertise), and Cross-departmental Coordination (data-silos prevent accurate model inputs). To overcome these, enterprises should: 1) Implement a centralized Risk Management Information System (RMIS) to collect real-time disruption data. 2) Partner with specialized consultants like Winners Consulting to bridge the technical expertise gap. 3) Adopt a phased approach—starting with high-impact scenarios (e.g., earthquake or cyberattack) before scaling to the entire organization. The priority should be establishing a data-gathering mechanism within the first 30 days, followed by model calibration within 90 days.

Winners Consulting Services Co., Ltd. specializes in Semi-Markov modelling for Taiwan enterprises, delivering compliant management systems within 90 days. Free consultation: https://winners.com.tw/contact