Studies have shown that product or service failures, such as a damaged parcel delivery or a delayed flight, are becoming more and more frequent (CCMC, 2020). As a result, customers either take their business elsewhere, complain to family and friends, or file a complaint with the company. When filing a complaint, customers can use different channels (e.g., Facebook, Twitter, telephone), with private (email) communication being one of the preferred channel choices for companies and consumers alike. It is essential for companies to carefully consider how they reply to such complaints as this affects future customer behaviour (e.g., loyalty) and profitability (Van Vaerenbergh & Orsingher, 2016). Studies have shown, however, that companies often do not deal with complaints effectively (e.g., by showing a lack of interpersonal care or not giving sufficient compensation) resulting in dissatisfied customers (CCMC, 2020).
The aim of this dissertation is to address two research gaps in the service recovery literature: a focus on language and discourse (e.g., how are responses to customer complaints phrased?) and on digital complaint handling (e.g., research is often limited to spoken interactions). These gaps will be addressed in five separate studies by paying attention to the discursive and linguistic features of the genre of email responses to customer complaints (ERC) and contextual parameters influencing (the effectiveness of) these features.
In Chapter 1, we examine the discourse structure and the linguistic characteristics of the ERC genre. The results show that this genre sets itself apart from related business communication genres (e.g., social media responses) as it consists of a unique blend of interpersonal (i.e., relationship-oriented) and transactional (i.e., content-oriented) strategies. Companies thus seem to feel safer to address the transactional dimension in complaint handling via email than via social media.
In Chapter 2, we examine the cross-cultural differences in communicative style between English and German ERCs by analysing their discourse structure and the frequency of first-person references (“I” and “we” pronouns). The differences we found indicate that English speakers tend to use a more interpersonal communicative style, while German speakers show a preference for a more transactional approach, which is in line with previous research (House, 2006).
In Chapter 3, we investigate the (de)legitimation strategies used by companies and customers in private and public computer-mediated complaint communication. Results show how companies differ in their approach to seek legitimacy in responding to customer complaints in email (private) versus on Facebook and Twitter (public). Customers, in their turn, delegitimate these companies by primarily criticizing companies’ service recovery failures and efforts. The results also reveal that social media agents lack agency to build relationships with customers and perform adequate service recovery.
In Chapter 4, we examine the role of interpersonal strategies in a business-to-business instead of a business-to-consumer context. While textbooks commonly recommend the use of interpersonal strategies when writing complaint refusals, service recovery research points toward contextual differences in this regard. We use an authentic sample of complaint refusals from an intercultural B2B setting to show that the decontextualized recommendations in textbooks are not always applied in actual practice and that this lack of interpersonal attention need not be problematic.
In the final chapter, we investigate the effect of linguistic realizations of employee empathy on brand trust in ERCs and examine whether this effect is moderated by giving a compensation (Study 1) or by the customer’s acceptance of blame for the service failure (Study 2). The results show that linguistic realizations of employee empathy positively influence brand trust through higher perceived empathy and perceived quality of complaint handling. In addition, the impact of employee empathy is more effective in a low versus high compensation condition, but is not influenced by the acceptance of blame when a partial refund is offered.
This dissertation concludes with a summary of the main findings and suggestions for future research.