Exploring customers' responses to online service failure and recovery strategies during Covid‐19 pandemic: An actor–network theory perspective

Abstract

While the debate on online service failure and recovery strategies has been given considerable attention in the marketing and information systems literature, the evolving Covid‐19 pandemic has brought about new challenges both theoretically and empirically in the consumption landscape. To fully understand customers' responses to service failure during a crisis we asked 70 millennials from three European Countries—Italy, France, and the UK—to describe their responses to service failure during the Covid‐19 pandemic (30 completed a 4‐week diary and 40 completed a 4‐week qualitative survey). Drawing on phenomenological, constructivist, and hermeneutical approaches, and utilizing an actor–network theory perspective, the current study proposes a new framework for understanding customers' responses to online service failure and recovery strategies during the Covid‐19 pandemic. Conclusions highlight implications for theory, policy, and management practice through extending comprehensions of service failure recovery processes by examining how marketing policies generate different social impacts during a crisis situation which facilitate the achievement of customer satisfaction and positive outcomes.

1. INTRODUCTION

Research reveals that the evolving Covid‐19 pandemic represents an unprecedented and unanticipated set of challenging circumstances for several business sectors, including hospitality and tourism (Škare et al., 2021), airlines (Belhadi et al., 2020), education (Brammer & Clark, 2020), and the fashion industry (Baek & Oh, 2020). Companies initiated and implemented several innovative marketing activities to address the heightened anxiety of their customers during the first and second waves of the pandemic. The combined effects of the Covid‐19 outbreak have seen several fashion flagship retailers either disappear or struggle for survival. No company was adequately prepared to change their service provision, and the majority continued with their usual modus operandi. Scholz (2020) stated that Covid‐19 has not just brought about the need for change, it also points a way forward: as an engine of innovation. Scholz's (2020) report also found that e‐commerce, as a share of US retail sales, increased as much in 8 weeks as it had in the previous 5 years. The Covid‐19 pandemic has led fashion customers to be more conscious about the sustainability of the clothing industry (Rabimov, 2020). Research has found that during the global pandemic, 83% of UK consumers were in favor of clothing items designed to last longer and clothing that is more repairable (Statista, 2020a).

A stream of researchers and media outlets have largely documented the impact of the Covid‐19 pandemic on consumer buying behavior, and have identified the social barriers that have an impact on their everyday life, particularly digital forces (Bacq et al., 2020; BBC, 2020a; Beaunoyer et al., 2020; Blackburn et al., 2020; Seetharaman, 2020; Shah, 2020; Sheth, 2020). More specifically, researchers have demonstrated that the Covid‐19 pandemic is increasingly prompting consumers to undertake behavioral changes, such as shifting their consumption to online environments, under conditions of mandatory domesticity (Hennekam & Shymko, 2020; Sheth & Kellstadt, 2020; Szymkowiak et al., 2020). In recent months, online shopping has become the major focus in terms of consumer habits. While some providers and industries have adapted well by leveraging digital media, other industries have struggled due to the need to evoke customer emotions as part of the service experience (Seetharaman, 2020). The fashion industry has been significantly involved in social media activity to build long‐term relationships with customers (Henninger et al., 2017; Kim & Ko, 2012). Fashion providers are increasingly facing challenges in meeting demand and maintaining image quality and authenticity, prompting many fashion brands to balance their digital transformation and brand identity (He & Harris, 2020; Lay, 2018).

As fashion retailers look for innovative ideas and flexible ways to deliver sustainable customer services in the wake of rapidly changing market environments, service failure is becoming an increasingly important mode of customer experience created by social distancing and lockdown rules (Donthu & Gustafsson, 2020; Osborne, 2020; Sharman & Stephens, 2020). A unique characteristic of these failures fueled by the Covid‐19 crisis is the breakdown of service during the process of delivery and/or outcomes, thus creating the emergence of dissatisfied customers (Duffy et al., 2006). In the event of these failures, customers expect effective service recovery to address their negative experience (Bitner et al., 2000; McCollough et al., 2000). Previous research on service failure and recovery (SFR) strategies demonstrated variation in customers' explicit responses to recovery strategy types. Some researchers examined the impact of online service recovery strategies on customers' perceptions of companies (Azemi et al., 2019; Ringberg et al., 2007; Wirtz & Mattila, 2004).

In our view, consumers' responses to online SFR differ under conditions of “mandatory domesticity and social distancing rules” when faced with threats such as the Covid‐19 crisis and they tend to rely on contextual rationality (Hennekam & Shymko, 2020, p. 790). We argue that consumers' responses to service failure along with expected recovery may not be as sensitive to objective probabilistic values as they are to emotions (Gigerenzer, 2006; Szymkowiak et al., 2020).

To our knowledge, no previous study has investigated how consumers' responses to SFR strategies are affected by crises such as the Covid‐19 pandemic. Azemi et al. (2019) provided some valuable insights into the contextual and pluralistic stances of consumers during the entire online SFR process. They identified three categories of customer responses to online service failure based on the customers' social world. First, exigent customers acknowledge their faults; however, they hold that the service provider is responsible for SFR processes. Second, solutionist customers become frustrated with failures but trust that employees, deemed to be responsible individuals, will provide solutions to the failure. Third, impulsive customers initiate recovery and are satisfied with service failure explanations. Azemi et al. (2019) surmised that customers' perceptions of SFR are influenced by their social world; however, they did not directly address the assumption that consumers' perceptions are directly influenced by crisis situations, such as the Covid‐19 pandemic. Understanding consumers' perceptions of SFR strategies might lead to better relationships and the development of an effective marketing strategy associated with current and future crisis conditions. We, therefore, extend Azemi et al. (2019) study by investigating the effects of the Covid‐19 pandemic on consumers' responses to service failure and expected recoveries when faced with unprecedented and challenging circumstances. Given the ubiquity and pace of the global pandemic, particularly in terms of the changing behaviors of consumers, the need to understand customers' responses to SFR has become vital for researchers and practitioners alike. The aim of this paper, therefore, is to assess both the changing nature of consumers' behavior during the Covid‐19 pandemic in the UK, France, and Italy and, equally, to assess consumers' perceptions and experiences of online SFR in relation to a crisis situation. Finally, of central importance for the current paper is the broader question of whether the impact of Covid‐19 leads to the emergence of new online SFR processes. The results of this study might enable firms to adapt their SFR processes in a crisis situation to achieve customer satisfaction and positive outcomes.

The present study investigates online service failures and expected recoveries from an actor–network theory (ANT) perspective. The corollary of ANT is that society and technology cannot be conceptualized as ontologically separate (though interrelated) entities (Doolin & Lowe, 2002; Latour, 1994). In this paper, we address the following questions: How do customers in the fashion industry perceive online SFR during the current Covid‐19 crisis? and How might one explain customers' expectations of SFR under these conditions? Our paper is structured as follows. The next section reviews existing literature related to SFR and it is conceptually presented through ANT; the third section, Methodology, summarizes the data collection and data analysis procedures. In the next two sections, Results and Discussion, we present the results and then consider the emergent data in relation to existing literature and our conceptual framework. The last two sections highlight the main theoretical contributions and further research directions.

2. THEORETICAL UNDERPINNING

2.1. Failure recovery strategy

Despite the rapid increase in online services, the increased digitalization of services has created barriers and disruptions for groups of customers and businesses. The service industry has witnessed specific mass service disruptions when normal operations are significantly disrupted by circumstances, like Covid‐19, beyond a firm's control (Beirão et al., 2017; Finsterwalder & Kuppelwieser, 2020; Heinonen & Strandvik, 2020; Kabadayi et al., 2020). Changes to service delivery have had a combination of positive and negative effects (Tuzovic & Kabadayi, 2020) and the struggle to deliver contact‐intensive services with high efficiency on digital platforms has resulted in technology‐based service failures (Beirão et al., 2017; Heinonen & Strandvik, 2020; Kabadayi et al., 2020).

Past research on SFR primarily focused on three main streams. The first stream examined the critical effect of SFR strategies on a firm's service quality and performance (Baliga et al., 2020; Durvasula et al., 2000; Van der Heijden et al., 2013). These studies mainly addressed the economic and customer retention benefits of SFR. Johnston and Michel (2008) suggested that formal complaint management procedures, embedded in organizational structures, led to better financial performance. The findings of Tax and Brown (1998) explicitly showed that ineffective SFR management can cause the significant costs to firms, including the time and money invested to repeat service procedures, compensating customers, losing customers, and the likelihood of negative word of mouth. Knox and van Oest (2014) emphasized the costs and benefits associated with effective recovery for different recovery scenarios, indicating the returns can outweigh the costs by reducing potential customer churn.

The second stream focused on dual recovery strategies that partially employ customers in firm‐initiated SFR strategies (Giebelhausen et al., 2014; Ringberg et al., 2007; Umashankar et al., 2017; Yim et al., 2012). Drawing on the idea of rapport between customers and service employees, DeWitt and Brady (2003) argued that existing rapport increased postfailure customer satisfaction, decreased negative word of mouth, and did not increase customers' motivation to complain about poor service. Heidenreich et al. (2015) argued that customers' participation in recovery must match the level of cocreation during service delivery. Smith et al. (2009) argued that customers' direct involvement in SFR requires directly asking for input on how service failure should be addressed or altering the recovery process based on customer input. Dong et al. (2008) results indicated that customers experienced higher levels of clarity, service value, and satisfaction in cocreated recovery procedures.

The third stream focused on customers' motivations to cocreate a SFR self‐service process (Bitner et al., 2002; Collier et al., 2017; Dao & Theotokis, 2020). Dong et al. (2016) extended the examination of co‐created recovery into the context of self‐service technology (SST), which increases customers' autonomy and responsibility in SFR procedures. This led to a stream of studies examining the effect of SST on consumers' perceptions of a firm's service delivery. Meuter et al. (2000) identified that SST provides customers with information‐learning benefits; some firms encourage customers to take the initiative using SST to recover from service failures on their own. However, Zhu et al. (2013) pointed out that because customers are recovering independently of the assistance of company personnel, SST design properties are significantly important to reduce customers' potential divergence from SST in SFR situations.

The above reflects the importance of SFR processes; historically, however, most studies did not focus on SFR processes under challenging conditions, such as Covid‐19. Indeed, researchers increasingly view SFR processes as multifaceted phenomena, with antecedents based on multidimensional interactions between a firm and its customers. However, contradictory findings with regard to SFR processes persist. What is clear is that SFR is a complex phenomenon and Covid‐19 may precipitate further complexity, leading to customers' dissatisfaction and frustration when they rely mainly on technology during a crisis situation.

Notably, technology failures restrict customers' usage of services, creating more dissatisfaction if the failure is unexpected (Meuter et al., 2000; Tuzovic & Kabadayi, 2020). Existing literature demonstrates that technology failures are especially critical because customers expect to access online services with high convenience and control (Keeney, 1999; Ozuem et al., 2017; Wang et al., 2010). As such, customers become highly dissatisfied if online services do not deliver as expected. As well as technology failures, fashion providers have been subject to other key service failures, such as late deliveries, a time lag between returns and resending to customers, stock‐out or over‐stock, and deliveries in single solutions (McMaster et al., 2020). Covid‐19 has caused major disruptions to the global supply chain (Altman, 2020; Ivanov, 2020) adding more risks to supply chain management and service failures. Research suggests several concerns about buying clothing items online, such as product quality, incorrect sizes (Robertson, 2020; Schild, 2019), and package deliveries (Bischoff, 2020). These further increase perceptions of service failure severity.

Prentice et al. (2021) suggested that the impact of Covid‐19 entailed the need to address the well‐being and sustainability of key stakeholders at the micro and macro level, including customers, employees, and service organizations, as well as the functional management of services. Lockdown and declining economic conditions have caused individuals with lower incomes and individuals with reduced social contact to be more at risk from the negative psychological outcomes of Covid‐19 (Tull et al., 2020), such as anxiety or worries regarding their online orders. Consumers' negative responses to service failures have become significantly higher as a consequence of these factors. With fashion providers already pressured by cancellations of orders from global suppliers (Statista, 2020b) and customers, there is an increased desire among providers to maintain interpersonal relationships with customers to reduce the negative effects of customers' dissatisfaction shared online (Chen et al., 2018; Christodoulides et al., 2021; Esmark Jones et al., 2018; Sun et al., 2017; Umashankar et al., 2017).

Two of the specific and important elements of recovery strategies are the speed of recovery (Baker et al., 2008; Borah et al., 2020; Crisafulli & Singh, 2017) and recovery efforts made by providers in response to customer complaints (Fan & Niu, 2016; Maxham & Netemeyer, 2002). Along with consumer anxieties about Covid‐19, employees now have restricted access to effective workplace infrastructures and to networks of colleagues to collectively manage services. This contributes to the lack of effective service delivery (Dixon et al., 2020), particularly for providers who have to reconsider how they will deliver their service and approach online customers (Diebner et al., 2020; Lee Yohn, 2020; Marr, 2020), and consider whether customers will seek pathways to apply their own efforts to the service recovery (Yi & Gong, 2008; Zhu et al., 2013).

Perceptions of technology usage may be altered by Covid‐19‐induced confinement (Beaunoyer et al., 2020) and, specifically, by feelings of isolation if social support is not present within online systems. When a service failure occurs, technology has traditionally applied specific recovery procedures, which can cause difficulties if the provider is trying to provide personal support and assistance to customers (Van Birgelen et al., 2002). This can cause customers more difficulty if they have to handle tasks they are unfamiliar with (Gelbrich et al., 2020), including how to respond to service failure. The increase in provider services needing to be conducted online, in turn, increases the risk of customers losing access to key support and guidance that digital agents may be unable to provide. In digital environments, perceived human support in technology‐mediated services is a factor for customers (Larivière et al., 2017; Rafaeli et al., 2017). Applying a social presence on digital platforms increases the likelihood of customers identifying a social entity within the digital environment (Van Doorn et al., 2017). This generates an interpersonal and supportive digital environment (Longoni et al., 2019). Companies have to reinforce their interactivity with customers to compensate for missed offline shopping experiences that customers enjoyed before the Covid‐19 pandemic (Gonzalo et al., 2020). This would strengthen customers' perceptions that they are the core focus of the service experience (Berg et al., 2020). This in turn suggests the need for companies to emphasize not only the speed of their recovery strategies and efforts, but also their customers' experiences of these processes. The aim is typically to decrease the level of customer dissatisfaction with service failures that may occur.

2.2. Consumption landscape and online shopping experience

Recent studies conducted by Addo et al. (2020) and Akhtar et al. (2020) highlighted the effects of the threats of Covid‐19 on offline shopping, such as hesitance in consumers' shopping behavior in response to threats of contagious disease and government restrictions. However, in the fashion industry, which has an offline and online service presence, the pandemic led to increased demand for online delivery and experience. While the physical retail industry was negatively affected, increasing reliance on e‐commerce and digital experiences compelled the enhancement of digital experiences and digitally based SFR procedures.

Researchers have posited that retailers seek to distinguish themselves from other competitors by offering customers a different shopping experience and motivating them to return for future purchases (Hult et al., 2017, 2019; Katsikeas et al., 2016). Other studies have documented that customers are both sensitive and demanding in terms of the quality of service delivery due to the benefits offered by technology (Gong & Yi, 2018; Parasuraman & Grewal, 2000). The demand for online services sets providers major challenges in maintaining customers' satisfaction in terms of shopping experiences (Danaher et al., 2003; Degeratu et al., 2000). Customers in online environments are more affected by their satisfaction with the experience due to the “cognitive lock‐in” effect (Shankar et al., 2003). This makes them more satisfaction‐sensitive to a range of factors, including product or service choice (Campo & Breugelmans, 2015; Degeratu et al., 2000), transaction costs (Chintagunta et al., 2012; Shi et al., 2018; Shih, 2012), intangible elements of shopping experiences and consequences (Laroche et al., 2005), and price dispersion of shopping experiences (Zhuang et al., 2018).

Hult et al. (2019) reported that customers view online purchasing as a convenient and efficient shopping method, but they perceive it as risky because of the uncertainty about the reliability of the product and service. The nonphysical context of online purchasing, which physically separates customers from the product and supplier, creates worry regarding whether the quality and performance of the product will match the product description advertised online, and the possibility of late or failed deliveries (Chiu et al., 2014; Glik, 2007). Vakulenko et al. (2019) suggested that customers have a holistic perspective of online services and do not distinguish the different roles of departments in e‐commerce procedures. This means that consumers' evaluations and satisfaction are influenced by both their online service experience and the delivery service components, which are shaped by many actors of the provider's service procedures. Similarly, Zeithaml et al. (2002) found that online service quality affects satisfaction and purchasing. They argued that service quality is measured through multiple dimensions, such as site design, ease of use, and reliability, rather than through a single dimension or process. Their study seems to suggest that if any dimension of online service is disrupted or fails to meet expectations, it will cause disruption to customer satisfaction in terms of their whole online shopping experience, unless the provider responds effectively to the failure.

2.3. Examining SFR with ANT

According to ANT, the emerging computer‐mediated marketing environment comprises heterogenous collectives of consumers, technologies, machines and objectives (Doolin & Lowe, 2002; Ozuem et al., 2008). ANT can be used to explore the influential links between human and nonhuman actors (Bencherki, 2017; Latour, 1987, 2005; Law, 1992; Zoo et al., 2017) and their impact on each other (Holmqvist et al., 2020; MacMullin et al., 2020). ANT provides a sociotechnical perspective on service failures or other intangible objects (Sarker et al., 2006) from which interactions between technology and human processes can be examined (Islam et al., 2019; Shim & Shin, 2019). In ANT, actors include both social and technical entities without making distinctions between the two (Baines, 2017; Islam et al., 2019; Sismondo, 2009). Walsham (1997) pointed out that ANT examines the motivations and actions of groups of actors who form elements, linked by associations, of heterogeneous networks of aligned interests. This capability requires the interweaving of both human and nonhuman actors, such as technological artefacts (Walsham, 1997, p. 468). Increasing use of the online consumption landscape during the Covid‐19 crisis strengthens the dynamic relationship between humans and technology in the fashion industry. Epistemologically, using ANT in SFR research, we assume that nonhuman actors potentially possess agency over human actors. The epistemic aim of ANT is to infer that reality is assumed to be actively performed by various actors in a particular time and place (Booth et al., 2016; Cresswell et al., 2010).

Recent studies have demonstrated that high levels of social media usage, along with the reported news on Covid‐19, have sparked psychological distress (Chao et al., 2020; Garfin, 2020). Other studies have noted that online activity prompts individuals to critically evaluate risks and information during times of crisis (Garfin, Silver, et al., 2020; Li et al., 2019; Ranjit et al., 2020) to maintain a level of normality in their everyday activities. Some individuals have been motivated to seek platforms containing social networks (Bansler & Havn, 2004; Loiacono & McCoy, 2018; Seo et al., 2018) to maintain connections with friends and family under Covid‐19 conditions. Thus, various groups of individuals will generate different emotional responses and actions towards the same online communication channel (de Kervenoael et al., 2017). In the event of service failures, a negative experience with one communication channel can negatively impact a customer's perceptions of the service experience as a whole (Patten et al., 2020). This means that technology‐based failures, or other failures including delivery failure, and inefficient customer service will impact on providers' service quality. It might be argued that customer dissatisfaction has been higher or more severe during the Covid‐19 crisis due to a shifting alignment between actors in an unstable and precarious environment.

3. METHODOLOGY

3.1. Paradigm of inquiry

To fully understand the data collected for this study we utilized a phenomenological hermeneutical methodological approach. In addition, this phenomenological, constructivist, and hermeneutical study uses an abductive approach and theoretical sampling techniques to provide an experiential, problem‐solving‐orientated theoretical framework for comprehending the impact of Covid‐19 on millennials' perceptions of SFR in three European countries. Phenomenological approaches emphasize the pursuit and explication of truth through the description of phenomena as they display themselves to consciousness. For the phenomenologist, there is a return to lived experience through capturing existence as it is lived. “Experience is the performance in which for me, the experiencer, experienced being “is there” in propria persona, in person, and is there as what it is, with the whole content and the mode of being that experience itself, by the performance going on in its intentionality, attributes to it” (Husserl, 1969, p. 94).

Through engagement with the world, experience is identified through “perception, imagination, desire, thought and so on” (Woodruff Smith, 2007, p. 189). Fundamentally, the basis of understanding the development of knowledge involves “the lived experience in the lifeworld” (Howell, 2013, p. 62). Through horizons or the transient perspectives of other individuals, we comprehend self and others through empathy. People (including those being researched as well as the researcher) are caught up in preconceptions and horizons; consequently, analysis and understanding can never be completely free of these as well as past experiences (Ozuem, Willis, Howell, Helal et al., 2021), which provides the underpinning for comprehending interpretivism, constructivism and hermeneutical perspectives.

To facilitate this philosophical perspective and methodological approach we employed a grounded theory sampling technique which enhanced the idea of “givenness” and how this emerges through the research process. Theoretical sampling recognizes and facilitates transformations in the research process through evolutionary change. Theoretical sampling allows autonomy and liberation in data collection processes and encourages theory generation through comparison controls. The application of “theoretical control over choice of comparison groups is more difficult than simply collecting data from a preplanned set of groups, since choice requires continuous thought, analysis and search” (Glaser & Strauss, 1967, p. 52). Corbin and Strauss (1998) considered that to sample theoretically one progresses in an evolutionary fashion. Theoretical sampling allows each sample to build on the previous data and analysis; during the research process, theoretical sampling becomes specific as the theory emerges and evolves. Through the development and continuation of preunderstanding and being‐in‐the‐world, theoretical sampling incorporates past, present, and future perspectives as the research and researchers develop theory and practice through interacting and analysing data. In addition, ANT emphasizes that the human actor is the dominant factor in network relationships (Somerville, 1999) and perceives human beings and objects as entities (John, 2009) that influence others. Social structures are not created simply by human actions; both human and nonhuman entities affect and connect with each other in the formation of reality. In the context of this study, we explored the construction of reality in terms of online interaction through an investigation of the development of this reality through a hermeneutical analysis of qualitative data. The literature review informed the development of the research questions in terms of Seetharaman (2020), who identified the need to evoke customer emotions; and Lay (2018) and He and Harris (2020), regarding the balance between digital transformation and brand identity.

3.2. Data collection methods

In terms of methodological orientation and fit (Dubois & Gadde, 2002), we employed exploratory qualitative methods with the aim of obtaining detailed insights into the perceptions of service recovery experiences under situational crises, and within a millennial consumer context. We conducted two studies using self‐completion qualitative diaries and qualitative surveys. Wilson (2016) recommended diaries as a useful method to capture real‐time information to explore concepts in studies applying hermeneutic phenomenology, and for participants with location preferences to record their encounters. A draft of the research instrument was constructed in a joint virtual meeting with three team members from Italy, France, and the UK. Following the refinement of the draft by the team members, each partner was tasked with carrying out the allocated data collection.

Collectively, the data collection methods for both studies are well suited to study SFR strategy in crisis situations. In addition, analysis of qualitative data is heavily descriptive and is typically presented in thematically organized patterns developed from the analytical stages, as explained in Section 3.3.

3.3. Participants

Italy, France, and the UK were selected for several reasons. First, these countries provide unique settings to explore the effects of the Covid‐19 pandemic on SFR. As of May 2021, France had more than 5 million confirmed Covid‐19 cases, followed by the UK and Italy with more than 4 million cases each, making them the top three European countries with the most recorded cases across the whole of Europe (Stewart, 2021). Second, Italy, France, and the UK before the Covid‐19 pandemic had the highest e‐commerce value in Europe, which tripled following the increased demand for online shopping after the start of lockdown (Coppola, 2021; McKinsey & Company, 2021); clothing was the most desired category of online purchases (Gautier, 2021). Finally, Italy, France, and the UK are a sample of European countries with “high levels of advanced technological capability” and the study took a millennial‐centric approach (Ferri‐Reed, 2014, p. 13) in which millennials were conceptualized as virtually interactive on social platforms (Daniel et al., 2018).

Contradictions occur across the literature regarding what constitutes millennials' generational age range, leading to varying interpretations of this demographic cohort (Gurău, 2012; Luo et al., 2018). Some studies have argued that the classificatory age range of millennials is overly fluid and varies from one social context to another (Dimock, 2019; Tolani et al., 2020). A plethora of meaningful distinctions suggests that the birth years of millennials are between 1980 and 2000 (Dimock, 2019). Millennials are highly networked and constantly engaged in digital systems (Melović et al., 2021); they expect a seamless shopping experience, are the growth engine of the fashion industry (Clark, 2021; Gharzai et al., 2020; Helal et al., 2018; Ozuem, Willis, Howell, Lancaster, et al., 2021) and are experts at finding alternatives during service failures. These distinctions make millennials a fitting phenomenon of interest in understanding how they respond to service failures and recovery choices under service‐disruptive crises like Covid‐19. We approached a 108 individuals, however, 70 actively participated in the study. For both studies, we recruited millennials between 18 and 39 years of age from three European countries: France, Italy, and the UK (see Table 1).

Table 1.

Demographic information of participants

Country	Gender	Occupation	Age
United Kingdom	Female	University business and administration student	18
	Male	Programme manager	30
	Female	Fashion shop assistant	28
	Female	University finance student	25
	Female	University marketing student	20
	Female	MSc International Business	23
	Male	University marketing student	25
	Male	Technical manager	31
	Female	Social media coordinator	28
	Female	Administrator	26
	Female	Service engineer	26
	Female	University marketing student	26
	Male	Sports college student	25
	Male	University humanities student	25
	Male	Human resource assistant	29
	Female	MSc International Business	30
	Male	MSc International Business	30
	Male	University business and administration student	18
	Male	Marketing coordinator	25
	Female	MSc International Business	36
	Male	University human resource student	26
	Female	University language student	37
	Female	Shop assistant	28
	Male	University economics student	26
	Male	Customer service operator	39
France	Male	Product manager	25
	Female	Shop assistant	24
	Male	MSc Global Luxury student	39
	Female	MSc Global Luxury student	36
	Male	Product manager	26
	Female	Promotion assistant	25
	Female	MSc Fashion Management student	38
	Female	MSc Fashion Management student	25
	Female	MSc Fashion Management student	39
	Female	MSc Fashion Management student	30
	Male	MSc Fashion Management student	25
	Female	MSc Fashion Management student	25
	Female	Cosmetics product manager	31
	Female	MBA Global Luxury Management	26
	Female	International key account manager	26
	Female	Cosmetics product developer	32
	Female	MBA Luxury and Fashion Marketing	25
	Male	Fashion shop assistant	18
	Female	University marketing student	22
	Female	Fashion web designer	28
	Male	Product manager	25
	Female	Shop assistant	28
	Male	Software test consultant	37
Italy	Male	University business and administration student	25
	Female	Fashion marketing consultant	27
	Female	Fashion marketing teacher	33
	Male	University fashion marketing and business student	28
	Female	University humanistic science student	26
	Female	University art and entertainment student	25
	Female	University engineering PhD student	28
	Female	University business and administration student	26
	Male	University business and administration student	25
	Female	University fashion marketing and business student	18
	Female	University chemical science PhD student	28
	Female	University art and entertainment student	26
	Female	University business and administration student	30
	Female	University fashion marketing and business student	18
	Female	University business and administration student	29
	Female	University business and administration student	30
	Female	Performing art actor	19
	Female	University business and administration student	22
	Male	Banking officer	39
	Male	Risk control officer	37
	Female	MBA student	36
	Male	Community banking branch manager	39

3.3.1. Study 1

Study 1 was a qualitative diary study conducted in the UK, France, and Italy. Forty‐eight individuals were approached and asked if they would participate in the study. Of those approached, 30 individuals (10 UK, 10 France, and 10 Italy) wrote about their experiences on a daily basis for 4 weeks focussing on SFR strategies during the Covid‐19 crisis. The use of diaries allowed individuals to record their daily experiences within locations convenient to them. Pressure to assign time to other daily activities essential to the participants, placed on hold by Covid‐19 restrictions, reduced the perceived efficiency of adopting virtual in‐depth interviews that were also restricted by specific time allocations. This prompted the usage of a data recording method that could be used at the participants' convenience (Wilson, 2016).

Another advantage of the qualitative diary method is the ability to examine the recorded experiences of participants within a spontaneous context (Bolger et al., 2003). This enabled participants to record real‐time events they personally encountered without the interference of the researcher, other than distributing and sending weekly reminders, and collecting diary records (Williams et al., 2016). Diaries capture the “‘lived experiences' of everyday life that fill most of our waking time and occupy the vast majority of our conscious attention” (Wheeler & Reis, 1991, p. 340). Thus, applying the diary method gave the researchers access to individuals' everyday recorded experiences over a time period, allowing them to identify changes in emotions and different events that unfolded (Fuller et al., 2003) within a 4‐week period. The immediate recording of events that diaries facilitated reduced the potential for biased results and allowed the collection of data to occur in a natural context (Bolger et al., 2003; Hennekam & Shymko, 2020).

3.3.2. Study 2

In Study 2, we used a qualitative survey method to collect data from participants. Semistructured questionnaires were sent to the participants electronically, enabling them to complete the survey at a place of their convenience, which is a beneficial factor for time‐limited research or samples of individuals that are geographically dispersed (Braun et al., 2017). This technique was useful for data collection during a situation such as the Covid‐19 crisis. Forty participants (15 UK, 13 France, and 12 Italy) successfully completed the survey over 4 weeks. It comprised 16 open‐ended questions concerning perceptions and experiences of SFR during the Covid‐19 crisis. Participants were sent weekly reminders to regularly record their experiences relating to SFR during the Covid‐19 crisis. Unlike other surveys that follow quantitative measuring methods, qualitative surveys seek to harness qualitative research values and techniques that lead to differing frames of reference (Braun et al., 2020). Qualitative surveys contain open‐ended questions centered on a particular topic, diverging from predetermined responses of the type elicited via quantitative surveys (Braun et al., 2020). This allowed the participants to respond in their own words and to provide rich and complex accounts of the type typically sought by qualitative researchers (Braun & Clarke, 2013). This data collection method offers the opportunity to find a range of perspectives, experiences, and sense‐making from participants (Braun et al., 2017).

3.4. Data analysis

In relation to the theoretical sampling, data analysis was conducted using the qualitative data analysis approach proposed by Gioia et al. (2013), whereby three stages (orders) of analysis were conducted to generate major themes and subthemes (Kaur et al., 2019). The first stage involved the examination of the qualitative responses to identify primary codes; the second stage saw the generation of subthemes based on the primary codes and through coding theoretical understanding to generate implicit ideas from the explicit transcribed data. Data from the qualitative diaries and surveys were analysed using a systematic approach conducted in three stages, referred to as first‐order, second‐order, and aggregate dimensions (Gioia et al., 2013). First‐order concepts represented ideas emerging from the detailed responses in the diaries and surveys. We used NVivo, a qualitative data analysis tool, to organize the emerging categories. Combining responses from individuals with concepts found in extant literature led to the second‐order, which yielded nine conceptual categories. The final stage involved aggregating the related patterns across the second‐order categories resulting in the generation of a final set of core concepts representing customers' perceptions of SFR during the Covid‐19 crisis. These were: the customer experience, emotional response, technology as resilience, and social isolation. Figure 1 provides a guide to the data analysis process conducted for this study. For example, data coding identified the first‐order concept, “Clients cannot be left with the consequences of a company's failures,” which was categorized under the second‐order category “Service management during Covid‐19” (Figure 1).

Figure 1.

Data structure

The final concepts served as a basis for a framework, which, when combined with the primary data, led to the identification of three customer types: blanders, empathizers, and churners (BEC). These customer types are characterized by their distinctive responses to SFR during crisis situations. These three customer types, along with the four core concepts, were used to develop a BEC framework (Figure 2). Our goal here was to understand how the identified concepts constituted different elements of BEC (see Figure 2) (Altinay et al., 2014).

Figure 2.

Blanders, empathizers and churners framework

4. RESULTS

4.1. Customer experience

The perceived impact of service failure on customers is dependent on their evaluation of the recovery efforts of providers (Sengupta et al., 2015). However, despite providers' best efforts, service recovery may not always be implemented as effectively as expected. Unsuccessful service recovery often causes a “double‐deviation” (Basso & Pizzutti, 2016) resulting in customer dissatisfaction being intensified. In contrast, successful recovery creates a recovery paradox; a situation in which a customer thinks more highly of a provider than they did before the recovery (Maxham & Netemeyer, 2002). In recent months, in addition to the barriers to service recovery, Covid‐19 has had a global impact on businesses, forcing major adaptations in consumption and operational behaviors (Carracedo et al., 2020, Donthu & Gustafsson, 2020; Eggers, 2020; Pantano et al., 2020). This is explained well by a 22‐year‐old Italian university business and administration student:

They said that because of the increase of the number of orders during Covid‐19, delays occurred … even so, the client cannot be left with the consequences of a company's failures and companies should get more organised during a pandemic.

Retailers are aware that their recovery responses will impact customer satisfaction, thus, compelling them to adapt their capabilities and operations to handle unprecedented demand and pressure caused by Covid‐19 (Gordon et al., 2020; Sjödin et al., 2020; Wang et al., 2020). An early study by McCollough et al. (2000) suggested that companies that prevent service failures from occurring maintain better customer satisfaction than they might have by responding to failures. Under the economic strain of Covid‐19 lockdowns, several firms could not deliver their services to their full capacity. This important point was mentioned by a 23‐year‐old British MSc International Business student:

I had two Debenhams gift cards. I discovered you are only able to use one card at a time. I would need to make more purchases than I intended to use these gift cards. So, I lose money and the company profits from it.

This participant's experience reflects a common issue that customers have encountered with providers going into liquidation during the pandemic. Some media outlets reported issues with gift cards having a 50% purchase limit, meaning customers would have to spend more than originally intended to use the gift card (Peachy, 2020). Additionally, customers were unable to use shopping vouchers for online purchasing and were left with expired cards that could not be used (BBC, 2020b). Gift cards and coupons, with an expiry date, apply pressure on customers to use them, evoking a use‐it‐or‐lose‐it mentality (Danaher et al., 2015). This indicates that with monetary losses for firms and customers, providers need to pay more attention to customers' possible concerns to reduce potential dissatisfaction, as indicated by this 28‐year‐old British social media coordinator:

Before I contacted the company, they informed me there could be more delays due to the Covid‐19 situation, and sent extra updates to keep me informed. It's like they already knew what my concerns were.

With Covid‐19 causing mass uncertainty for individuals, providers face pressure to anticipate customers' concerns and monitor potential failures that may occur. An important customer belief, which firms must anticipate, is that firms could prevent service failures (Choi & Mattila, 2008). Customers are more likely to react harshly towards firms that caused them to endure negative outcomes, particularly those that could have been prevented (Grégoire & Fisher, 2008). Some perceive the provider as lacking the ability and desire to prevent or reduce service failures.

4.2. Emotional response

The Covid‐19 crisis has increased individuals' emotional responses, particularly to economic pressures, and the need to feel socially connected with organizations and other customers following the physical restriction policies (Karpen & Conduit, 2020). Several psychological factors have arisen from Covid‐19, including gratitude and empathy. These have encouraged individuals to reflect on themselves and their expectations (Kirk & Rifkin, 2020). In this study, empathy is a major customer response that has arisen following service failures in the context of Covid‐19. In the context of Covid‐19, many retail companies are facing extra operational costs and the challenge to adapt their service delivery to accommodate larger demand and customers' new needs (Pantano et al., 2020). This 24‐year‐old French shop assistant noted:

The tension of Covid‐19 makes it easier to get nervous and angry. But thinking about firms having to remain open and handle increased workloads, patience and empathy helped me get through late deliveries.

Similarly, this 26‐year‐old French product manager stated:

I don't feel like complaining about late deliveries, these days there are a lot of other problems to deal with.

These respondents noted that the pandemic is affecting business performance. Under non‐Covid‐19 circumstances, it is possible that customers would have been less forgiving if providers did not resolve or avoid failures. As regards service failures, customers are more likely to forgive failures that are beyond a firm's control, and less likely to forgive errors if they perceive the firm has the ability to avoid or manage failures (Klein & Dawar, 2004; Monga & John, 2008). The Covid‐19 crisis has highlighted the increased effect of customer emotions on online purchasing behavior. Two types of purchasing behavior are identified: impulsive and frugal purchases. Impulsive purchases are made without prior planning and are connected with emotional regulation (Fenton‐O'Creevy et al., 2018). In recent decades, interest in achieving happiness has increased due to the desire to improve overall well‐being (Etkin & Mogilner, 2016). In recent months, restrictions on outdoor social activities have increased people's need to be emotionally indulged, indirectly causing customers to lower their self‐control on purchasing behavior. Customers with low self‐control are happier to indulge without needing a reason (Petersen et al., 2018), identifying short‐term pleasure as their priority (Poynor & Haws, 2009). Such behavior is identified by several individuals, including this 19‐year‐old Italian performing art actor:

I am purchasing a lot more online than I did before Covid, as a kind of consolation to the fact that we are in lockdown and I am bored at home…and I am buying more for fun than necessity, to have some emotions.

The pandemic has caused major disruptions to social gatherings and activities like shopping; thus, many individuals seek activities that stimulate emotional experiences with little concern for an economic expense. In contrast, customers who conduct frugal purchasing worry if they are making a “smart” choice (Hampson et al., 2018). This means that customers will act on rational thinking in relation to their purchasing, rather than acting on pure impulses. Covid‐19 has particularly increased the importance of frugality and the need to understand frugal customers (Mehta et al., 2020). Many customers may have been impacted by the economic consequences of the pandemic, as indicated by this 20‐year‐old British university marketing student:

I need to be more attentive because when you buy online it is difficult to know about the quality of a fashion product … right now I have to be careful with money since getting less working hour revenue due to Covid‐19.

Intriguingly, this respondent indicates an economic concern for online purchasing under Covid‐19 pandemic. Frugality, they suggested, is not limited to making low‐price decisions, but decisions that provide long‐term satisfaction, fulfillment, and happiness to individuals rather than short‐term impulses (Mehta et al., 2020). Frugal shoppers may search for items based on the idea of “good value for money,” meaning that a combination of cost, quality, and even personal preference will be considered. Thus, frugal customers may have an emotional response to online shopping, but will apply more self‐control to their purchasing as opposed to impulsive customers (Petersen et al., 2018).

4.3. Technology as resilience

Prior studies have examined the perceived importance of technology for customers, particularly the ability of technology to empower individuals and their control over online activities (Eisingerich et al., 2019). Technology helps individuals achieve focal goals like satisfaction and personal growth (Wolf et al., 2020). Indeed, since the pandemic, there has been an increase in the usage of the internet for entertainment purposes (Beech, 2020) and customers are relying more on technology to continue service activities like healthcare, education, and shopping (Wyman, 2020). Additionally, there has been a rise in usage of social media as a method of communication as seen in the rise in online customer engagement with news media, the entertainment industry (Arens, 2020) and the fashion industry (Berg et al., 2020). Given that technology enables convenient and immediate message exchange (Gelbrich et al., 2020), many customers rely on social media to communicate with providers, as suggested by this 38‐year‐old French MSc Fashion Management student:

These days, I do more online shopping and socialisation than I did before. To contact customer service, I use Vestiaire Collective's Facebook chat, you don't have to wait long to hear from their staff.

Technology is not just used as a method to conduct services or activities, but as a method of providing support to individuals undergoing stressful situations (Gelbrich et al., 2020). With the majority of services being conducted online, and restrictions on social meetings the “new normal,” customers are becoming reliant on human support incorporated into technology services to cope with stressful situations imposed by the pandemic. This was certainly the experience of this 30‐year‐old British MSc International Business student:

Since the pandemic, social media has brought people closer to cope with the situation… having staff communicating information and support through social media helped me cope with service failures I encountered.

Interestingly, this respondent indicates the importance of instrumental and emotional support; both are strategies aimed at helping individuals cope with negative situations (Duhachek, 2005). Instrumental support provides objective change through advice about what to do, whereas emotional support involves improving one's mental state (Carver et al., 1989). Easy‐to‐use services and customers' emotional well‐being are both important to providers. Support and assistance through digital services are especially important when customers need to complete challenging tasks (Gelbrich et al., 2020); support can help customers overcome or cope with barriers inflicted by crises such as the pandemic.

However, customers increased usage of technology, fueled by the Covid‐19 pandemic, has had an impact on the performance of many businesses. While big brands and larger firms with existing digital systems have adapted quickly, others firms, particularly small firms or firms with limited digital services, have struggled to maintain efficient performance (Iansiti & Richards, 2020). The demand for more digital services has motivated companies to invest in digital development (Venkataraman et al., 2020). Some have used social media to reach their audience, but have struggled to deliver efficient quality due to limited technological tools (Iansiti & Richards, 2020). This French 28‐year‐old fashion web designer noted:

Not all firms have technological knowhow. Small stores I know tried using Facebook, but it wasn't easy to communicate with someone.

Similarly, another participant, an Italian 27‐year‐old fashion marketing consultant, commented on a particular firm's difficulty in adapting their digital services stating:

A store called Privalia, their site doesn't work from a phone or tablet, that's how I normally shop. With Privalia's poor site, I don't believe that using it to shop or communicate a complaint would be easy.

These comments suggest that firms face issues in maintaining the sufficient quality of their web‐based services.

4.4. Social isolation

Loneliness and social isolation are two of the predominant outcomes of Covid‐19 (He & Harris, 2020). Some consumers become vulnerable under these circumstances; some endure an economic loss following a service failure or have limited understanding of how businesses are operating under the pandemic. Individuals in vulnerable situations may turn to online service providers to receive social support (Fletcher‐Brown et al., 2020); researchers have stressed the need for providers to adapt their operating standards and place individuals' well‐being at the center of their values (Donthu & Gustafsson, 2020; Sigala, 2020). Recent research suggests that during Covid‐19 pandemic, customers are more concerned than usual with how service problems are solved (Dao & Theotokis, 2020). This 31‐year‐old British technical manager commented on their experience of a provider's response to a failure:

I get an email saying I need to pay an extra €10 to return and replace the broken shoes. This made me feel worse and like the company was trying to push me off them. I had enough and stopped using that company, they lose customers' trust by profiteering from their failure.

With many individuals having to cope with isolation and customer service complexities, customers are more likely to be negatively impacted by recovery approaches they perceive to be inadequate given the type of service failure. The pandemic has prompted society to be more empathetic to others undergoing difficulties whilst being socially isolated. However, before the pandemic, research examined how customers perceived brands that mistreat customers (e.g., Folkes et al., 1987; Lepthien et al., 2017). Some customers may think that providers should aim to deliver recovery strategies that benefit customers the most, regardless of the pandemic, as suggested by this French 22‐year‐old university marketing student:

It doesn't matter if we are under a pandemic, no company should push their customers away by expecting them to pay more for a disservice.

This respondent indicated that customers can compare a company's prepandemic behavior with their behavior during Covid‐19 crisis. Some customers use SST to continue purchasing, but may be unable to fix failures themselves and require employee assistance (Dao & Theotokis, 2020). A provider's initiative to offer extra support can positively impact customers' judgments, regardless of the outcomes (Dao & Theotokis, 2020), as indicated by the experiences of the following participants:

An 18‐year‐old Italian university fashion marketing and business student noted:

With Zalando, they went beyond considering the pandemic crises. I was able to talk to a service operator and was compensated. I never felt like I had to solve the issue myself.

A 33‐year‐old Italian fashion marketing teacher noted:

Under pandemic, you have no choice but to rely on the chatbot. You feel more anxious about service failures, so it's great when companies provide help.

An important factor raised in these responses is the perception that these customers did not experience the consequences of service failure by themselves, and could rely on the providers for support. Research suggests that customers appreciate recovery that emphasizes sympathy and concern compared to economic recovery which is perceived as transactional and emotionally detached (Wei et al., 2020).

5. DISCUSSION

This study draws on a number of relevant concepts in formulating our conceptual framework. As abductive research, we combined both deductive and inductive approaches. Emergent data were inherently iterative and recursive (Klag & Langley, 2013). Following the data gathering and after the initial data analysis, we moved back and forth between the emergent data, themes, dimensions, concepts, and the relevant literature (Gioia et al., 2013). Following Gioia et al.'s line of reasoning, we crystallized both the knowing and not knowing elements to capture the deep embeddedness of the participants' views in theoretical terms, and, ultimately, our understandings of existing theory. This process allows for the discovery of the findings that describe our phenomena of interest. Our findings suggest that the essential processes of online recovery strategy during crisis situations can be meaningfully categorized into three customer types: blanders, empathisers, and churners.

Blanders are loyal customers in terms of purchasing from their brand preferences, but when service failures occur, they expect normal, if not exceptional, recovery strategies that are fair to customers during crisis situations. Although customers may acknowledge that the situation or failures may have been beyond a provider's control (Klein & Dawar, 2004; Monga & John, 2008), blanders are not willing to forgive if the provider does not continue to retain their satisfaction and trust. In other words, blanders are likely to perceive a crisis situation as an unjustified reason for a company's delivery of poor customer service. Exigent customers, one of the three customer groups identified by Azemi et al. (2019), seek immediate recovery results and have expectations of the provider and the SFR process. With a high degree of sensitivity towards the provider, exigent customers will seek revenge if a service recovery paradox is not achieved. In contrast, this study describes blanders as committed customers who seek a positive resolution in light of the crisis situation they are under. Regardless of the crisis, they will continue to use the services of their preferred brands using technology to maintain access to these services. Though they may experience feelings of betrayal if there is unsatisfactory service, they typically employ positive coping strategies, such as seeking compensation or recovery solutions, which are aimed at reaching a resolution with the brand. This contrasts with purely seeking revenge (Weitzl & Hutzinger, 2019) as Azemi et al. (2019) exigent customers would seek if their expectations were not met. Thus, blanders have high expectations of providers, but will try to maintain a lasting relationship with the brand if the brand is willing to overcome service barriers.

Empathizers, like blanders, have a positive relationship with their favorite brands and will seek to maintain this relationship with the brand even if it was directly responsible for a service failure (Cheng et al., 2012; Sinha & Lu, 2016). They harbor negative perceptions regarding service failures, but they are far more forgiving than blanders and will respond more positively to any recovery attempt providers make, underlining the “love is blind” effect (Grégoire & Fisher, 2008; Weitzl & Hutzinger, 2019). A key characteristic of empathizers is attitudinal loyalty; this loyalty towards the brand causes reviews or complaints from other customers to have less influence on their decision to remain with the brand (Langan et al., 2017) and reduces their dissatisfaction with service failures.

Azemi et al. (2019) solutionist customers trust that providers will deliver the necessary recovery procedures, but their trust in the provider is conditional on a successful SFR delivery and they will hold the provider responsible as exigent customers do. In contrast, empathizers consider the impact of the Covid‐19 crisis on a company's ability to deliver customer service. Customers may often empathize with employees and firms by considering their perspectives and intentions during a service interaction (Bove, 2019; Wieseke et al., 2012). This empathy is more likely to cause customers to recognize and be more sensitive to the actions taken by others; thus, causing them to become less concentrated on the negative issues (Davis & Oathout, 1987; Pera et al., 2019). Empathizers thus consider how a crisis may be affecting a brand and its personnel and are more likely to reflect on the firm's efforts to deliver services despite the crisis. Additionally, empathizing customers may serve as social support for a firm and their personnel, such as by providing suggestions, encouragement and assurance to companies (Bove, 2019; Ngo et al., 2020) through the companies' service or social media channels.

Churners, in comparison to blanders and empathizers, have limited to no emotional connection with brands, and reflect characteristics linked to behavioral loyalty, which consists of repeated purchase behavior without having a psychological attachment to a brand (Gorlier & Michel, 2020). Azemi et al. (2019) third customer group, impulsive customers, are defined as being more patient with service failures or with services they have little experience of in comparison to exigent and solutionist customers, who respond through negative electronic word of mouth (eWOM) following successful recovery. However, impulsive customers are prone to switching to other brands if advised by close relations despite recovery success (Azemi et al., 2019). Although the churners in this study share behaviors similar to those of impulsive customers towards brand relationships, churners' mindset is opposite to that of impulsive customers and blanders and emphasizers towards SFR evaluations. Churners focussed on the service failure, irrespective of the brand or the crisis situation. Research suggests that although customers may be disappointed in an experience, they might not take it as a personal loss, and are more prone to switch to another brand than invest in reconciling with the brand following recovery offerings (Tan et al., 2021). This means that churners are more attentive to how the provider approaches and resolves service failure and are less concerned about repairing or starting a committed customer relationship with the brand; this makes them unresponsive to companies' weak or strong recovery efforts during crises. This also means that churners pay more attention to the experience than the brand. Research suggests that brand disappointment is related to the actual poor experience that follows service that fails to meet individuals' expectations (Zeelenberg & Pieters, 1999). Churners will be more likely to evaluate the experience, which will influence how they perceive the brand. For example, if a firm fails to deliver a service following a customer purchase and if it does not provide compensation that meets the customer's expectations, then the brand may be perceived as profiteering. The likelihood of customers becoming churners increases when customers have more brand options (Amin et al., 2019); for example, another brand may have an improved reputation for delivering services during a crisis situation than the brand the churners have previously or recently used. Thus, churners may search for providers online with specific criteria in mind, such as how a firm has adapted to the external crises, as well as price and service convenience.

6. CONCLUSION

The findings of the present study show that, given the specific context of Covid‐19, consumers' responses to SFR are emotionally mediated based on a challenging consumption landscape.

The findings identify distinct emotional responses from customers that providers may encounter during crises, such as the current global pandemic. This study suggests that customers evaluate SFR procedures during crises on the basis of their psychological attachment to a brand. Customers with less experience of a specific brand are less determined to seek a resolution from the provider or consider how their actions may impact the provider, compared to customers who have a connection with and experience of a provider. Additionally, a crisis, such as the Covid‐19 global pandemic, may influence customers' sentiments towards life and consumption behavior until the situation improves. Just as positive brand relationships and psychological loyalty characteristics can influence customers to remain with a brand despite the marketing efforts of other brands (Thomson, 2006; Umashankar et al., 2017), customers' attachment to brands can overcome the influence of crisis situations on their perceptions of a brand's recovery procedures. Yet, loyal customers' evaluation of service recoveries will differ depending on how they incorporate the circumstances of a crisis situation. Customers with no attachment to a brand may be indecisive regarding their perception of service recovery and loyalty to brands, as they are motivated by habitual purchasing or monetary benefits (Gorlier & Michel, 2020). This calls for providers to adapt their recovery approaches, particularly on compensation and customer support, by enhancing their focus on customer orientation support services.

This study identifies three types of customers. Empathizers are motivated to identify the best in what providers deliver regardless of the consequences. A group of customers who have high expectations of providers during a crisis are called blanders. The final category, churners, switch to brands that deliver high purchasing convenience and monetary returns regardless of the brand's recovery effort. Both empathizers and blanders demonstrate a positive attitudinal loyalty towards their favorite brands, but respond differently to firms' management of their services during crises. Empathiser customers pave the way towards the practice of “business as usual” and they repeat their patronage with preferred brands despite any difficulties encountered before or after service recovery. Any recovery procedure delivered by brands enhances empathisers' gratitude to the brand, as their evaluation reflects on how a crisis may be affecting the provider's operations. In contrast, blanders express the need to be provided with normal or superior customer service, accentuating the point that providers must prioritize their customers. Blanders will probably judge the severity of a service failure to be higher if the provider does not deliver an efficient recovery response. However, unlike churners, blanders seek to maintain their relationship with brands through service recovery reconciliation. For churners, providers must provide monetary recovery strategies, such as prompt return of money, rewards and promotions.

When dealing with blander customers, providers must employ both monetary and social support recovery methods. Though blanders have an attachment towards specific brands, providers must consider the possibility that blanders will spread negative eWOM. While some customers turn to the provider to resolve service failures, some immediately turn to social media to vent their frustration against the brand (Grégoire et al., 2015). Thus, it is important for firms to quickly identify dissatisfied customers and initiate engagement through social media and involve other customers in the service recovery communication procedures. Customers within online communities act as jurors towards viral eWOM related to brands and can settle in favor of the brand if brand sentiment is strong (Ibrahim et al., 2017). However, providers should encourage these customers to become a source of support for other customers rather than simply be sharers of information. Social support from social relationships is as important as informational support and can increase the perceived credibility of brand endorsements (Errmann et al., 2019). The social support offered may reinforce a customer's positive sentiment towards the brand, allowing the firm the opportunity to deliver recovery procedures with less likelihood of negative eWOM intensifying the process. Empathisers have a strong emotional attachment to a brand, which reduces the likelihood of them engaging in negative eWOM and increases the likelihood that they will offer support and assurance of the brand's quality (Dost et al., 2019), including the brand's recovery efforts and outcomes. Considering the impact empathiser customers have on providing social support and suggestions to firms and their personnel (Bove, 2019; Ngo et al., 2020), it is recommended that fashion brands use empathiser customers to deliver emotional support and positive eWOM to other customers like blanders.

This study is the first to empirically investigate how customers respond to SFR processes during a global pandemic crisis. Thus, it extends knowledge about customers' emotional reactions towards firms' service recovery processes. The study identifies how customers responses to SFR strategies in a crisis are more emotionally inclined and incorporate the crisis environment and their relationship with brands. The characteristics and behavior of customers described in this study may provide insight to marketers on how customers respond to crises and what recovery strategies would be most appropriate for them under such circumstances. To maintain customer retention, providers must focus their attention on their customers' state of mind and demonstrate genuine care towards them. While customers appreciate the challenges firms face from the global pandemic, it is important to deliver customer service that is perceived by customers as treating them with fairness and dignity. Such treatment will increase the likelihood of customers' forgiveness for service failure, even if the service failure was not fully resolved (Weitzl & Hutzinger, 2019). Given the impact the crisis has on firms financially, customers may not receive a full recovery from service failures. Thus, it is recommended that governments, in times of global crises, consider developing policies that reduce or protect customers from the severe negative effects of service failures during crisis situations.

7. FURTHER RESEARCH DIRECTIONS

Our framework should be interpreted in light of the given crisis situation. Although the current study focused on the global pandemic, it is important to note that other crises, of different natures in different countries, may occur and influence customers' behaviors. Furthermore, the responses of the European customers we described in this study were influenced by individualistic cultural behaviors which may differ from the collective cultural mindsets found in other countries. The generalizability of the BEC framework developed from this study could be examined if future research tests the framework under different crisis contexts in other countries. The study focused on a global pandemic, yet the findings were collected from three countries located within Europe. Customers' responses to service recovery during a crisis situation could be examined in a worldwide context, thus extending not just the number of countries but also the variety of countries from different continents. For studies examining crises on a global scale, quantitative methods are recommended to gather and examine the volume of data from the proposed larger sample.

Ozuem, W., Ranfagni, S., Willis, M., Rovai, S., & Howell, K. (2021). Exploring customers' responses to online service failure and recovery strategies during Covid‐19 pandemic: An actor–network theory perspective. Psychol Mark, 38, 1440–1459. 10.1002/mar.21527

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.