Analysis of Voice of Customer data through data and text mining

Abstract

The importance of great customer service to building a strong business has prompted increased interest in applying text analytics to customer data. To demonstrate power data and text-mining techniques, we present two case studies outlining the use of advanced text analytics for the analysis of Voice of Customer (VoC) data collected from both external and internal customers of Taco Bell Corporation. External VoC data was collected through several channels over three years and analysed using text-mining techniques. Using historical data, a detailed taxonomy of keywords that typically occur in customer comments was developed to characterise these comments into meaningful categories, with sentiment scoring rules applied to the keywords for further classification. Over 2,000,000 customer contacts were analysed, and the findings were correlated to the structured data collected on the surveys, providing key insights on product, service and facility topics. The impact on overall satisfaction was measured for each topic area, providing focus for the operation of the restaurant. A second case study highlights how data and text-mining techniques provided actionable insights, allowing the internal information technology support service desk to implement changes that improve call answering times and reduce the impact of these issues on sales transaction revenue.