In a world of ever-growing customer data, businesses are required to have a clear line of sight into what their customers think about the business, its products, people and how it treats them. Insight into these critical areas for a business will aid in the development of a robust customer experience strategy and in turn drive loyalty and recommendations to others by their customers. It is key for business to access and mine their customer data to drive a modern customer experience. This article investigates the use of a text mining approach to aid sentiment analysis in the pursuit of understanding what customers are saying about products, services and interactions with a business. This is commonly known as Voice of the Customer (VOC) data and it is key to unlocking customer sentiment. The authors analyse the relationship between unstructured customer sentiment in the form of verbatim feedback and structured data in the form of user review ratings or satisfaction ratings to explore the question of whether customers say what they really think when given the opportunity to provide free text feedback as opposed to how they rate a product on a scale of one to five. Using various Sentiment Analysis approaches, the authors assign a sentiment score to a piece of verbatim feedback and then categorise it as positive, negative, or neutral. Using this normalised sentiment score, they compare it to the corresponding rating score and investigate the potential business insights. The results obtained indicate that a business cannot rely solely on a standalone single metric as a source of truth regarding customer experience. There is a significant difference between the customer ratings score and the sentiment of their corresponding review of the product. The authors propose that it is imperative that a business supplements their customer feedback scores with a robust sentiment analysis strategy.
|Number of pages||27|
|Journal||International Journal of Data Warehousing and Mining (IJDWM)|
|Publication status||Published (in print/issue) - 1 Oct 2019|
- Machine Learning
- Sentiment Analysis
- Text Analysis
- Text Mining