خدمات هوشمند سیستم مدیریت کیفیت بر اساس تجزیه و تحلیل و پیش بینی صدای مشتری

This study suggests the intelligent service quality management system to analyze the causes and effects of VOC (Voice of Customer) variation and to forecast its occurrence based on the former study in the service industry, especially insurance company. In the competitive business environments, where customers are considered as a key success factor of the company, our research will help the company achieve the pro-activeness towards VOC and improve the quality of customer service based on scientific grounds. The proposed system is designed with three phases: the filtering phase to detect significant variations, the pattern detection phase to generate VOC occurrence patterns, and the VOC forecasting phase. In the filtering step, VOC is calculated and normalized to get rid of apparent exaggeration. In the pattern detection step, internal factors such as product or service qualities are used as sources for generating regular patterns and external factors like sales policy and customer inflow are used as sources for generation of irregular patterns. At the last phase, we forecast VOC based on the pre-defined pattern of VOC occurrence. We evaluate the proposed methodology by applying to the real VOC in a life insurance company.

The service industry company offers intangible products and services to customers. Intangibility makes the quality of service changeable; service can be varied by agents, economical environments, and even customers themselves. In the service industry under these uncontrollable factors, customer complaints occur more frequently and more diversely than in the manufacturing industry. In today’s challenging business environment, VOC is considered to be critical to the business in identifying problems and giving opportunities for all parts of companies. However, companies do not know how to achieve these goals and VOC managers are still locked in the tedious and time consuming data collection and reporting, which fails in improving problems. In many insurance companies, we found that the studies on the occurrence and variation of VOC depend on the knowledge from experts’ experiences not on systematic and scientific method. This extemporaneous approach may derive wrong results and even if the results are correct, these are mere personal know – how not the knowledge stored in the knowledgebase that can be transmitted in the company. Furthermore, they try to find what causes the VOC variation after it has occurred, so it is difficult to confront the VOC occurrence proactively and quickly. For preparing, resolving sudden VOC, and preventing VOC eventually, the company needs an intelligent system that indicates right improvement points and forecasts VOC variation beforehand. With this system, the company will reduce the cost of staffing in dealing VOC and will improve the service quality.

In this paper, we proposed the VOC analysis and forecasting system. VOC includes various customer requirements on product or service, so roles a bridge to link customers and corporate. To detect patterns on VOC occurrence, the proposed methodology focuses on environmental factors to reflect the characteristics of insurance industry as well as internal factors such as product or service quality. We examined the relationship between VOC occurrence and the pre-examined factors, and then we forecasted the VOC occurrence pattern based on their relationships. We applied our methodology to a Korean insurance company and tested the accuracy of forecasting method based on real VOC. We proved the validity of our proposed model and gained the managerial implications. Also, we implemented the proposed methodology as a web-based system. For a further study, the insurance agent needs to be examined as external factors even though we could not include it because of its complexness. Insurance agents have lots of uncontrollable characteristics; further study will systemize the characteristics of agents and derive controllable information based on the analysis on their characteristics and VOC occurrence. Also, we will extend our research by considering customer surveys on customer satisfaction as a former indicator of the change of VOC occurrence. Latent VOC can be detected by inquiring to customers directly and VOC occurs with a time gap of the time when customers feel complaints and the time when customers express it. Through this research, we will gain more information of customer requirements and improve the VOC forecast.