A Data Quality Framework for Customer Relationship Analytics

Poor data quality has become an increasingly pervasive problem for organizations leading to operational inefficiency, increased costs, and missed opportunities. As high quality data is a prerequisite to trusted data analysis, we propose a framework that focuses on improving the data model to improve data quality. In particular, we show how changes to the underlying data design can achieve key data quality properties. We conduct a case study that demonstrates the application of the framework to a customer relationship management (CRM) problem. Our evaluation shows that a set of CRM queries can be efficiently run over data sizes of up to 10 million records, and organizations can glean new insights about customer preferences and activity.