In today's competitive environment, it is becoming a standard practice to analyze data to predict market trends and to improve enterprise performance. However, in order to be truly putting customers at the heart of everything, we present a framework which looks at predictive end-user application trigger. This paper discusses current issues of business intelligence systems and then outlines our vision of future predictive end-user application trigger. The research aims to formulate a generic (high-level) prediction model that is able to capture, analyze and model the heterogeneity of customer's profile in terms of Web searches, calling patterns, calendar entries, emailusage, connectivity, contexts and device preference in order to predict services that would be pertinent to a specific user at a specific location and time. This paper talks about a number of examples of applying these technologies to improve BT's systems and services. © 2008 IEEE.