The use of visualizations to access and process enterprise data has been proposed as a boost to the usability of enterprise systems. Yet, novel user interfaces do not always yield the expected gains in performance. To realize the potential benefits from using a visual approach, it is important to examine the effectiveness of visualizations in a specific task context and to follow an iterative, evaluation-driven process for identifying and addressing shortcomings in their design and execution. In this paper, we present the results of a study examining the effectiveness of a visual interface called Association Map (AM) for exploring ternary associations of data versus a typical tabular interface from a popular enterprise system. This latest version of AM evolved from earlier versions that underwent similar evaluations and we discuss the design choices that were important for achieving user performance gains in accuracy and time-on-task compared to earlier versions and to table-based interfaces. Findings presented here help build the case for incorporating task-driven visual interfaces into traditional table-based representations used in enterprise systems to improve both the user experience and task-related outcomes. Furthermore, they provide insights for investigating the theoretical foundations behind what makes a visual interface work.