Not all data are created equal": Inequality in utility and implications for data management

With the rapid growth of organizational data resources and associated costs, understanding the effect of data management decisions on cost-benefit trade-offs becomes critical. This study links these trade-offs to the utility contribution of data resources. The magnitude of inequality in utility, the extent to which data records differ in their business-value contribution, is shown to be a fundamental driver of key data management decisions involving data quality management, system and database design, data acquisition and retention policies, and pricing of information products. By adapting and developing statistical tools for modeling and measuring inequality, this study analyzes some typical utility distributions and demonstrates the possible effects of inequality on utility-cost trade-offs and the overall net-benefit. A low magnitude of inequality is shown to result in "all-or-nothing" decisions. High inequality is likely to involve a more refined treatment of data resources based on their utility contribution, and will possibly require the development of management policies that differentiate these resources accordingly.