Despite significant technological progress, predictive analytics is still rarely used in a professional way for financial forecasting. An integral approach which goes beyond the mere technical application is necessary for a successful implementation. By the analysis of different implementation projects, different success factors for predictive analytics projects could be derived and summed up in an integral concept. Based on that improvement in the areas of forecast accuracy, the use of resources and steering effects can be realized.