Overview and details of the sessions of this conference. Please select a date or location to show only sessions at that day or location. Please select a single session for detailed view (with abstracts and downloads if available).
New Dynamics of Customer Complaints on Social Media: Introducing Data Mining Methods
Authors: Thomas Reimer (University of Rostock, Germany), Banu Aysolmaz (University of Maastricht)
When customers are dissatisfied with an organization or its service, they increasingly express complaints via social media. The viralness and possibility for follow-up comments increases the reach and impact of customer complaints. Thus, company’s social media efforts are becoming increasingly important.
In this paper we aim to gain a better understanding of the process of complaint recovery and factors that contribute to the recovery satisfaction. We address three major research questions: (1) What is the influence of the changed conversation structure due to the interference of other customers. Specifically, the influence of unaffected other users that virtually interact with the complainant, i.e. providing help, defending the company or confirming the complaint. (2) To what extent does recovery speed influences the satisfaction with recovery performance? (3) How to optimize the activity selection and process sequence of an online complaint conversation to satisfy complainers?
To address the research question (1), the Social Influence Theory (SIT) is used to explain how complainants are influenced by the presence and behavior of virtual others. In service recovery, virtual presence and interaction should enhance a complainant’s emotional and behavioral responses. To answer research question (2) we make use of the justice theory, assuming that fast responding to customers’ complaints can reduce customer anger and uncertainty and signals that the company cares about their customers’ comments. In order to address research question (3), we have used an innovative research approach. For the structured analysis of a large social media data set, we have used data mining methods to identify weak points or preferred sequences in the complaint handling.
We analyzed 1000 complaint-related twitter conversations of a big Chinese smartphone producer. The company's recovery performance was coded as positive or negative, depending on whether the complainant responded that the problem had been satisfactorily solved or still existed without an updated solution. We used the well-known process mining software Disco.
Our preliminary analysis shows some interesting findings that can be used to cope with customer complaints. The results show that SIT is applicable to the service recovery context in virtual environments. The interference of other customers decreased the satisfaction with the recovery handling both due to follow-up complaints and answers by brand advocates. Which shows that customers prefer the complaint handling just with their company. Furthermore, firm’s high response speed and low number of events to achieve a solution is a crucial indicator for service recovery satisfaction. Finally, process path analyses revealed that there exist favorable sequences, but also activity pairs which should be avoided. Especially, loops and repetitions decreased the satisfaction with the complaint handling. Thus recommendations for companies can be derived how to cope with customer complaints and how to take into account virtual interactions between customers.