In modern computer architectures, access latency varies considerably between different levels in the memory hierarchy. Consequently, applications with data access patterns that don't reuse much data in fast levels of the hierarchy incur additional delays. While most performance tools associate metrics with functions or statements, we explore data-centric analyses that associate metrics not only with data accesses but also with data objects themselves. Our tool focuses on attributing metrics to different variables and recording all the uses in the calling context. With very low overhead, the tool demonstrates its unique capability in identifying poor locality in various applications.