When decision support systems are implemented in practice, a number of special aspects and requirements must often be considered. From an operations research (OR) perspective many of these can be very challenging and/or are counterproductive for the stated objective. They can appear in any part of the optimization process that includes problem formulation, modelling, solution and result evaluation. In this paper, we describe and discuss a number of aspects and requirements in industrial implementations where optimization has been a vital planning or execution tool. We discuss the reasons behind the requirements, the OR challenge, how the requirements were approached and implemented and present results and general experience acquired from these implementations. The applications cover tactical, operational and real-time applications for several areas including transportation, scheduling, routing, inventory and process control.