Allocation of resources to improve security is crucial when we consider people’s safety on transport systems. We show how a system engineering methodology can be used to link business intelligence and railway specifics toward better value for money. A model is proposed to determine a probability of a success in service management. The forecasting model is a basic Markov Chain. A use case demonstrates a way to align statistical data (crime on stations) and probability of investment into resources (people, security measures, time).