This study proposes an information architecture and an accompanying data analytics model that together enable manufacturing companies effectively utilize their data for making tactical and strategical decisions. The main challenge in linking data and decision making is the extant inconsistencies and lack of compatibility between information architectures and the decision making mechanisms in many corporate. Our proposed holistic approach aims to gap the bridge between data science and proactive decision making. Based on the proposed architecture and model, we propose a practical roadmap for manufacturers to become more proactive in their business development and capacity planning. We specifically focus on the maintenance, repair and overhaul (MRO) systems, and demonstrate and validate our approach with a real life application via our ongoing collaboration with a major aviation MRO company. Maintenance-repair-overhaul (MRO) service providers are constantly subject to internal and external challenges due to unsteady workload, congestion, stringent customer requirements, etc. As workload fluctuates over time, work teams can speed up or speed down to cope with fulfilling deadlines and other customer requirements. As we illustrate using a case study from the aviation MRO industry, the relationship between workload and the service rates is typically nonlinear. Although data analytics can be employed to capture this relationship, incorporating it into decision making on planning and cost management is a nontrivial task. Collecting such data in right format and transforming them into useful input for predictive models require a multi-faceted information architecture. Subsequently, the output of the predictive models must be fed into decision making models. In this case study, we demonstrate how such a mapping from data to decision making can be designed and implemented effectively. The proposed holistic methodology assigns service orders to time slots based on the workload-throughput mapping. Depending on the current and future scheduled workload, the decision module aims to minimize the total cost of delaying orders and proposes optimal release (start) times for arriving service requests. The solution algorithm proposed for the decision making module first identifies an initial feasible solution by formulating all turnaround times for upcoming orders based on their original request dates. Then, it makes gradual improvements based on a survival analysis. We demonstrate that the proposed approach can provide useful valuation of service time slots for revenue management.