Service failure handling and resilience amongst airlines in Nigeria

Abstract

Organizations build competitive advantage via managing, and learning from, service failure cases. Proficient complaint handling depends on the operator’s strategic position to timely anticipate and respond to customer dynamics in terms of what interests them amidst disgusts. Therefore, this study extends research and theories on complaint management by proposing and testing a theory-driven conceptual framework that captures the proactive initiatives to deal with consumer dynamics. Data collection spans 403 participants in a field survey questionnaire and analysis involved SEM via AMOS 5.0. The results showed that empathy was most critical covariance with agility, followed by compensation and effort effects with anticipatory ability; attentive, facilitation and effort effects with adaptability; and facilitation, attentive and compensation effects with agility. Disgust passengers prefer more of symbolic than utilitarian recovery strategies; thus, we recommend one-on-one marketing activity and hybrid recovery package as critical for reinstating the disgusts.

Keywords:

• service failure
• complaint handling
• resilience
• post-recovery satisfaction

PUBLIC INTEREST STATEMENT

Organizations build competitive advantage when they proficiently manage service failure cases, and learn from them. This is because effective complaint redress is subject to the operator’s ability to timely anticipate and respond to customer dynamics in terms of what interests them amidst disgusts. Therefore, this study extends inquiries and theories on complaint management by proposing and testing a theory-driven framework that captures the proactive initiatives to deal with consumer dynamics. Data were collected from 403 participants via survey; analysis involved structural equation modeling (SEM); and the software was AMOS 5.0. The results showed that empathy had the most critical covariance with agility, followed by compensation and effort effects with anticipatory ability; attentive, facilitation and effort effects with adaptability; and facilitation, attentive and compensation effects with agility. This suggests that disgusted passengers prefer more of symbolic than utilitarian recovery strategies; thus, we recommend one-on-one marketing activity and hybrid recovery package as critical for reinstating the disgusts.

1. Introduction

The airlines make almost a humongous contribution to the global economy and the aviation industry in particular. Cooper (2017) posits that the socio-economic growth, especially in the knowledge economy, is intertwined with the developments in the aviation industry of which the airline operators play flagship role. Report from the Nigeriafinder.com shows that there are 2, 235 airlines competing globally, 345 operate in Africa, and 14 local and 29 international carriers operate in Nigeria. Similarly, Oxford Economics (2017) proposed that annually, over 2billion passengers and nearly 50 million tons of freights fly globally with a contribution of over 3.4 USD trillion, which is much more than the GDP from pharmaceutical, textile and automotive industries. The North American airlines, which are in the forefront with a net profit of about 13 USD billion as of 2015 against 200 USD million for African airlines, were projected to make 8.5 percent net-profit in 2017 (IATA, 2016). Further, FAAN (2017) and IATA (2016) assert that airlines in Nigeria create different categories of employments of about 700, 000, and generated about 8.2 USD billion or less than 0.5 percent to the GDP through the contribution declined in 2017 owing to reduced number of aircrafts that embark and/or disembark in Nigerian airports. However, the key performance indicators (KPIs) of the Nigerian airlines seem worst in the global and African scenes, given Daramola’s (2014) posit that South Africa, Kenya, Ethiopia, and India generate at least 2 percent of their GDP from airlines.

These suggest that the airlines provide a market with huge industrialization potentials. Significantly, the airlines drive political and socio-economic independence and encourage the exploitation of comparative advantage to increase foreign earnings, globalization, flexibility and innovations. In Nigeria FAAN (2017) recognizes that for almost a decade now, many airlines have operational problems that precipitate cutbacks on service provisions and in worst scenario, retraction and divestiture. Outputs fall to about 5 percent between 2015 and 2016 (FAAN, 2017), given the decline in domestic flight operations to 33 percent in the first quarter of 2017. Similarly, issues of safety and customer-service delivery by the Nigerian airlines seem abysmal; for instance, between 2003 and 2008, there were 11 air disasters claiming 335 lives and other major air crashes between 2010 and 2012 (Daramola, 2014). Scholars (Bamford & Xystouri, 2005; Tiernan et al., 2008) posit that passengers are exposed to ugly cases of abuse of rights, poor service packs, bags and baggage delays, lost/damaged luggage, and indiscriminate flight delays and cancellations due to weather and/or technical faults. Consumer Protection Unit (CPU) of Nigeria Civil Aviation Authority (NCAA) affirms that of the 43,196 flight schedules by eight domestic airlines from January to September 2016, 24,075 schedules were delayed; whereas 854 were out-rightly cancelled. Similarly in the first quarter of 2017, the News Agency of Nigeria (NAN) reported 6, 789 flight delays and 318 outright cancellations out of 10, 366 local flight schedules. Daramola (2014) attributes these ordeals, amongst others, to poor maintenance culture of the aircrafts, lack of sufficient technical experts, debt burdens and economic downturns, stringent government policies, dearth of aviation fuel and supply chain disruption, and weak government protection against stronger foreign and well-established competitors.

Aside from these narratives implying that service delivery in the airline operation is a complex one, given that too many avoidable and unavoidable factors could cause low-quality services; they urgently call for cross-context inquiries involving Nigeria and other developing countries. Similarly, the narratives suggest that the consequences of service breakdown are supposedly direr in Nigeria, and could cause providers to critically learn from consumer dynamics and post-consumption experiences in attempt to reposition trust and minimize the detrimental effects of service mishaps (Balaji et al., 2017; Cai & Qu, 2018; Cantor & Li, 2018; Tang et al., 2017). Amidst the airlines’ socio-economic position and the peculiarity of their surging service breakdowns in Nigeria, extant airline studies (Bamford & Xystouri, 2005; Chen & Chang, 2010; Rhoades & Waguespack, 2008; Sim et al., 2010; Tiernan et al., 2008; Waguespack et al., 2007) and others (Mostafa et al., 2014; Othman et al., 2014; Park et al., 2014; Tsai et al., 2014; K. Wang et al., 2014) on service recovery undoubtedly interest active providers and academics. These studies predominantly emphasize classification schemes (Abd Rashid et al., 2014; Awa et al., 2016); and provide correlational or anecdotal supports on the effect of recovery on customer satisfaction and post-complaint behavior (Abd Rashid et al., 2014; Kau & Loh, 2006; Sharifi et al., 2017). They span exploratory, scenario-based experiments and least generalizability, and the outcome of their theoretical and empirical results appears contradictory (Petnji et al., 2013) and under-estimates that the high degree of user-developer personal contacts, within which likely misunderstandings may result in failures.

Further, the inquiries least recognized the implied position of R. Lee and Lee (2013) that subject to the environment, customer’s animosities may be temporally unstable. Arguably, if the influence of complaint handling on satisfaction and post-complaint behaviour is indeed temporally unstable; then, the generalizability of results of the growing single-shot of aforementioned cross-sectional studies and more becomes questionable because of their mixed and inconclusive results that explain the differences in the prevailing environment. Again, previous studies (Hua, 2012; Vaerenbergh et al., 2014; Xu et al., 2014; Zhou et al., 2013) highlight the significance of recovery instruments and steps required for effective handling of disgust experiences, without a direct link to resilience even though resilience critically unlocks and/or sustains strategic windows. The confirmed criticality of recovery instruments implies linking them to measures of resilience in order to build competitive advantage (Kau & Loh, 2006; Maxham & Netemeyer, 2002; Michel et al., 2009), given that the comparative effects of the alternative failure-handling strategies have cross-context undertone. The relatively unclear effects of recovery instruments raise scholarly worries, especially in the context of airline services; and linking the instruments to resilience informs the currency of academic discourse, and the need to anticipate and timely adapt to customer dynamics.

Therefore, this study extends research and theories on complaint management by proposing a theory-driven conceptual framework that captures utilitarian and symbolic virtues, as well as proactive initiatives to deal with customer dynamics when handling service failures. First, this objective is justified by the scholarly (Maxham & Netemeyer, 2002; Othman et al., 2014; Tang et al., 2017) calls on regular examination and extension of extant recovery studies amidst the cross-sectional and/or environmental contexts. Second, credence to this objective is based on weak generalization of extant studies; the need to know if customers of different services and environments display similar behaviour; and the need to know if our findings will glean from similar complaint handling studies in other contexts. Finally, the objective is further reinforced by the need to grasp the direction of passengers’ ideal recovery package, amidst divergent studies. Whereas some studies (Bamford & Xystouri, 2005; McDougall & Levesque, 1999; Wirtz & Mattila, 2004) emphasize utilitarian recovery as most critical; others (Osarenkhoe & Komunda, 2012; Mostafa et al., 2014; Othman et al., 2014) propose symbolic and psychological recovery. The paper is structured to reflect a review and proposition of theoretical and conceptual frameworks, research approach and data collection, testing the conceptual framework, discussion, conclusion and recommendations.

2. Conceptual framework and hypotheses development

Service failure defines service performance or product-delivery attributes falling short of the customer’s ideal expectations; it includes out-of-stock of a service, slow and errors in delivery, unpleasant aroma, unbearable noise and waiting time, and inconsistent service (Istanbulluoglu et al., 2017; Othman et al., 2014; Vaerenbergh et al., 2014). If left unfixed, they pose serious threats to the provider, given that they create customer dissatisfaction and intention to switch (Balaji et al., 2017; Diaz et al., 2017), break down user-developer relationship, and generate negative publicity (Ha & Jang, 2009; Zeithaml et al., 2006). However, service failure handling is an exchange process, where customers report their mishap experiences, and the developer attempts to reinstate them via recovery programmes. The airlines are besieged with preventable and unpreventable service issues—the preventable issues are service interruptions occasioned by industrial strikes (Bamford & Xystouri, 2005), mishandled baggage (Rhoades & Waguespack, 2008; Sim et al., 2010), overbooking (Chan, 2000) and employees’ attitude to complainants (Chen & Chang, 2010); and the unpreventable issues mainly involve flight delays and cancellation owing to technical defects and weather conditions (Tiernan et al., 2008; Waguespack et al., 2007). The unpreventable issues are inescapable to every carrier and less amendable to marketing and building of competitive advantage. Scholars (Abd Rashid et al., 2014; Istanbulluoglu et al., 2017; Tsarenko & Strizhakova, 2013) posit that even amongst the best providers, service mishaps and complaints are unavoidable due to human and non-human errors, as well as the unique nature of services and the inseparability of production and consumption; thus, firms seek antecedents of expectancy disconfirmation and customer displeasure. Similarly, recovery paradox supports the practical inevitability of service-errors and suggests reinstating the affected customers in a manner at least equals their ordeals (Balaji et al., 2017; Michel et al., 2009).

Experience shows that the various “moments of truth” in service encounters may still be vulnerable to breakdowns; thus, scholars (Awa et al., 2016; Vazquez et al., 2010) represent complaint handling as critical intention to correct problems, to improve product quality and to potentially turn dissatisfied customers into satisfied. The exercise defines operator’s effort to identify and address perceived errors; to limit their harms, re-establish reputation and promote retention; to reposition trust during the service encounter before complaints and after the encounter when something had gone wrong; and to dissuade sharing of negative experiences, litigation and sanction by consumer activists and consumer right organizations (Battaglia et al., 2012; Del-Río-Lanza et al., 2013; Smith et al., 2012). Providers proficiently anticipate, prevent and recover service-failures since it costs more to attract (than to retain) customers in a competitive market, and customers themselves prefer on-going and event-driven accords to switching (Awa et al., 2016; Istanbulluoglu et al., 2017). Similarly, providers lose customer lifetime value when service fails, given that approximately 90 percent of dissatisfied customers avoid the provider (Business Week, 1984) and shares their experiences with 10 to 20 others whereas satisfied customers infest only 4 or 5 individuals (Abd Rashid et al., 2014; Osarenkhoe & Komunda, 2012). Other studies (Kau & Loh, 2006; Michel et al., 2009) show that customers need as many as 12 positive recovery experiences to overcome the negative effects of one ugly experience. Often dealing with disgust experiences involves utilitarian and pecuniary (compensations, refunds, discounts, or replacements), as well as symbolic and psychological (apology, status, respect, esteem, or empathy) dimensions of social exchange theory (Abd Rashid et al., 2014; Smith et al., 2012).

Supposedly, as markets for airline services evolve into more competitive and dynamic scenes, players attempt to reposition resilience and customer-endorsed services to enhance immediate and cumulative satisfaction amidst mishap. Edmondson (2011) posits that competitive advantage suffices when operators genuinely develop creative culture of learning from service mistakes, and attempt to transform complaints into service innovation. This informs why the complaint handling team vigorously seeks out, deals with, and learns from the problems even when unreported. Osarenkhoe and Komunda (2012) affirm that proficient recoveries are critical for enhanced customer satisfaction and customer engagement, and for preventing customer defection. However, many customers still remain unsatisfied despite the efforts by players to improve their services through managing complaints effectively. Again extant frameworks on complaints management from either the financial sector (De Matos et al., 2013; K. Wang et al., 2014); mobile telephony and telecommunications (Kau & Loh, 2006; Mostafa et al., 2014); online retailers (E. Lee & Park, 2010; Y. Wang et al., 2011); airlines (Chan, 2000; Chang & Chang, 2010) or hotels and/or restaurants (Othman et al., 2014; Park et al., 2014) have shown weak cross-context generalizability. This paper develops a hybrid conceptual framework that has utilitarian and symbolic values to the disgusted passengers, and the capability to proficiently accommodate customer dynamics in order to capture opportunities and turn stronger by using experiences to engage, adjust and adapt programmes that recover disgust experiences.

Scholars propose apology (Davidow, 2003; Wang & Mattila, 2011); acceptance of responsibility for service failure (Blodgett et al., 1997; Tax et al., 1998); prompt response (Davidow, 2003; Hua, 2012); staff empowerment and correction (Chung-Herrera et al., 2010; Seawright et al., 2008); attentiveness and explanation (Vaerenbergh et al., 2014; Xu et al., 2014); effort and empathy (Awa et al., 2016; Gruber & Frugone, 2011); facilitation (Bhandari et al., 2007; Casado et al., 2011); compensation (Davidow, 2003; Park et al., 2014); and follow-up in writing (Osarenkhoe & Komunda, 2012; Lewis & Spyrakopoulos, 2001) as alternative complaint handling options. The conceptual framework captures attentiveness, empathy, efforts, compensation, and facilitation as the dimensions of the independent variable (IV) because apart from their theoretical and empirical supports, as well as their utilitarian and symbolic virtues; they are more amenable to the airline setting than the other instruments. Airline transaction carries high involvement and high risks, given that scholars posit that the personal relevance of their services (Awa et al., 2016; Schiffman & Kanuk, 2009), criticality and severity of failures (Hess, 2008; Park et al., 2014), timeliness (Davidow, 2003; Seawright et al., 2008), and the existing developer-user social interactions (Gruber & Frugone, 2011; Ha & Jang, 2009; Zeithaml et al., 2006) are paramount for decision-making. However because consumers aggressively seek information and complain about disgust experiences that are expensive, infrequent, risky, expressive and ego-involving; a hybrid of utilitarian and symbolic driven conceptual framework is quite critical (Figure 1). The hybrid values of recovery packages will extend and/or complement the work of Bamford and Xystouri (2005), which suggested free business class tickets to any destination, printed letters of apology signed by the president of the airline, food vouchers, free hotel accommodations and re-booked flights free of charge as most common service recovery strategies.

Figure 1. The proposed conceptual framework

Correlating these IV dimensions with resilience gives the framework a further scholarly boost because apart from majority of extant studies (Ashill et al., 2005; Osarenkhoe & Komunda, 2012; Mostafa et al., 2014; Othman et al., 2014) relating the IVs to customer satisfaction, repurchase intention, and word-of-mouth publicity; every airline wants continuity, stability, competitive advantage, profitability and shareholders’ values through resilience, and managing tomorrow from today. Though resilience may be implied by such extant studies, establishing facts through scientific investigation is germane for effective decision-making, especially in the developing economies where the least of such inquiries is done. Resilience is a mechanism that assists organizations to formulate strategies billed to identify customers’ changing contexts, to take advantage of change, and to improve upon current states (Conboy, 2009; Hamel & Valikangas, 2003). It represents variation-selection-retention (VSR) mechanism, an evolutionary theory which, according to Weick and Sutcliffe (2007), emphasizes inherent characteristics to respond quicker, recover faster, or develop more uncommon ways to do things under perturbations than others. Amidst the huge external threats in the aviation industry, airlines that adopt resilience would ordinarily have the capacity to bounce back, to maintain positive adjustment, and to succeed amidst challenges and diversities (Conboy, 2009; Olsson et al., 2004; Shin et al., 2012). However, since the aviation industry is an all-around high involvement industry, resilience is critical for players and provides a strong basis for sustainability amidst customer dynamics and other environmental changes.

Kantur (2015) proposed four properties of resilience—agility, adaptability, robustness, and anticipatory ability—though the strength of robustness to recovery actions was not captured here, because Conboy (2009) and Zamenopoulos and Alexiou (2007) posit that every organization retains uniqueness in attributes and maintains core processes and configurations in times of changing business scenarios. Agility explains change readiness potentials and preparedness to speedily respond to surprises and mitigate possible threats (Conboy, 2009; Volberda, 1996; Walker et al., 2004); anticipatory ability conceptualizes the sensing of customer dynamics, aggressive prediction of future surprises, and making of solution-based and real-time decisions ahead of time (Zamenopoulos & Alexiou, 2007); and adaptability relates to learning to innovate by altering strategies, operations, and structures in order to cope with disruptions (Olsson et al., 2004; Shin et al., 2012; Walker et al., 2004).

2.1. Empathy

Empathy represents psychological therapy, feelings or apology of some sort, aimed at dousing the anger of an aggrieved. It is the provider’s admission of mistakes and acknowledgement of the complainant’s ordeals and acceptance of responsibility to reinstate him to his ideal state(s) if they (the ordeals) are legitimate. For instance, “we are sorry for what happened, we will make-up and we assure you it won’t reoccur” are symptoms of a cost-free calm-down process that expresses the provider’s concern for the consumption-related ordeals. Barlow and Moller (1996) posit that empathy uses the expression of understanding, concern, politeness, courtesy and respect to avert customer’s guilt. A good use of empathy encompasses working with customer dynamics and making decisions ahead of time in order to combine and reconfigure efforts to take advantage of change. Studies suggest that empathy programmes are developed with agility and flexibility to foresee and manage customer challenges (Hamel & Valikangas, 2003; Kantur, 2015; Volberda, 1996); are regularly adjusted to deal with the dynamic environment (Chu, 2015; Walker et al., 2004; Zamenopoulos & Alexiou, 2007); and rules and processes modified to cope with the environmental challenges (Olsson et al., 2004; Shin et al., 2012; Tapanainen, 2012). We hypothesize below:

H1: There is no significant relationship between providers’ empathy in times of service mishaps

and their agility to embrace the complainants’ environment.

H2:There is no significant relationship between providers’ empathy in times of service failures

and their anticipatory ability to embrace the complainants’ environment.

H3:There is no significant relationship between providers’ empathy to disgusted customer and

their ability to adjust to the complainants’ environment.

2.2. Facilitation

This explains the ease with which an injured consumer accesses providers, registers his complaints, and perhaps receives hassle-free, friendly and timely resolutions. Dissatisfied consumers are encouraged to voice out their complaints when the providers anticipate and prepare ahead of complaints clear policies, rules, structures, and procedures that simplify guidelines for filing complaints and access to frontline staff (Ashill et al., 2005; Davidow, 2003; Mostafa et al., 2014). Scholars (Balaji et al., 2017; Edmondson, 2011; Hess, 2008) observe that encouraging dissatisfied customers to complain and making complaint-mechanism simplified impact critically on resilience and the likelihood to complain. The theory of empowerment relates very much to flexibility in that amidst facilitation; employees are tactical and flexible in handling complaints in order not to miss resilience (Othman et al., 2014; Seawright et al., 2008) and/or avoid what McColl-Kennedy and Spark (2003) referred to as neutrality (same treatments to all) since no two human beings are same. Organizations develop speedy, tactical and flexible approach to facilitation because they have proactive capabilities to foresee customer challenges (Conboy, 2009; Tapanainen, 2012); they have anticipatory ability to be reworked overtime to accommodate dynamically complex environment (Walker et al., 2004; Zamenopoulos & Alexiou, 2007); and they are continuously refined and have modified roles, processes and structural relationships to cope with the environment (Shin et al., 2012; Stocia et al., 2003). Other studies (Conboy, 2009; Volberda, 1996) suggest that the indicators of facilitation are aided by firm’s ability to learn from change and to generate varieties of responses to unexpected and/or unfamiliar changes. Therefore, we hypothesize as follows:

H4: There exists no significant relationship between facilitation that proficiently handles

customer ordeals and provider’s agility to embrace the complainants’ environment.

H5: There exists no significant relationship between facilitation that proficiently addresses

customer ordeals and the provider’s anticipatory ability to embrace the complainants’

environment.

H6:There exists no significant relationship between facilitation that proficiently deals with

customer ordeals and the provider’s ability to adjust to the complainants’ environment.

2.3. Compensation

Compensation represents redress and key customer-endorsed recovery element that involves putting the wrongs right or re-establishing equity. The premise is that the complainers must at least be returned to status-quo (if not more) otherwise they remain dissatisfied with the response packs. Scholars (Casado et al., 2011; Kansal & Singh, 2013) posit that firms make fair policies, procedures, and rules for pecuniary and non-pecuniary compensations, including service replacements, free-services and service-upgrades, free-ancillary services, repairs and corrections, coupons, and refunds as well as no charges for filing complaints and minimal costs (if any) to obtain the package. An effective compensation package is that which recognizes the environmental dynamics-customers’ idiosyncrasies; given that certain types or mix of redress may affect customer behaviour differently. The need to understand and respond to the environment spans resilience and consideration of the three basic principles of redress—equity, equality, and needs (Davidow, 2003). For instance, if a flight is cancelled, then every passenger is entitled to receive full refunds (equity); passengers of all class receive the same treatment regardless of what they paid for the ticket (equality), and if there is only one seat space in an alternative airline, then it goes to the passenger that needs it most irrespective of what was paid, or how similar passengers were treated (need). Scholars (Hamel & Valikangas, 2003; Kantur, 2015; Walker et al., 2004) affirm that statistical criticality between compensation and resilience; they found that the design of compensation must be capability of accommodating customer dynamics; and disruption in customer demand and its pace must be effectively managed to capture opportunities by using experiences to engage, adjust and adapt programmes that will recover disgusted consumers.

H7:There is no significant relationship between compensation redress in times of service

mishaps and the provider’s agility to embrace the complainants’ environment.

H8: There is no significant relationship between compensation redress in times of service

failures and the provider’s anticipatory ability to embrace the complainants’ environment.

H9: There is no significant relationship between compensation redress in times of service

failures and the provider’s ability to adjust to the complainants’ environment.

2.4. Effort

Effort to accomplish recovery means front-line officers exerting power, energy, or bustle to psychologically deal with the mishaps (Kansal & Singh, 2013; Vaerenbergh et al., 2014) through the expression of concern, courtesy, respect, and regrets (Battaglia et al., 2012; Xu et al., 2014). In recovery context, effort represents the proportion of positive commitment used up by the staff to compensate the complainants. Arguably, performance rises when front-line staff show commitment and give much effort to get customers’ ordeals sorted out successfully. Lewis and McCann (2004) report that the commitment and effort to redress complaints affect resilience and repurchase rates—46 percent of complainants want money refunds (pecuniary), 25 percent want complementary deals, and 29 percent want proper response spanning more pleasant relationships (non-pecuniary). Studies show that the recovery team makes quick adaptation to changes and ensures goodness of fits with the recovery effort (Chu, 2015; Stocia et al., 2003); their efforts continuously refine and modify, and rearrange roles, processes and structural relationships in order to cope with the challenging environmental demands (Tapanainen, 2012; Walker et al., 2004).

H10: There exists no significant relationship between recovery efforts that generate pleasurable

outcome and the provider’s agility to understand the complainants’ environment.

H11:There exists no significant relationship between recovery efforts that generate pleasurable

outcome and the provider’s anticipatory ability to predict the complainants’ environment.

H12:There exists no significant relationship between recovery efforts that generate pleasurable

outcome and the provider’s ability to adjust and adapt to the complainants’ environment.

2.5. Attentiveness

Attentiveness is the personalized psychological care and attention billed to calm disgusts and to address angers, emotions and disappointments (Awa et al., 2016; Battaglia et al., 2012; Davidow, 2003). Attentiveness uses courtesy, personalized responses, and willingness to listen to address complainant’s feelings and reasons for the complaints (Davidow, 2003; Vaerenbergh et al., 2014). Scholars (de-Rio-Lanza et al., 2009; Kim et al., 2009) assert that the forms and contents of employees’ response influence resilience and customer satisfaction since it is not only failure to deliver the core services that causes dissatisfaction but also the employees’ manner of response to the failures (double deviation). When handled with good knowledge of the environment, attentiveness is a single most important element of service recovery, having the largest effect on resilience, customer satisfaction and repurchase intention (Davidow, 2003; Xu et al., 2014). A strategic mix of attentiveness has the capacity to predict and forecast likely opportunities and threats, discount their effects on operations, and propose possible alternative actions and reactions to deal with ordeals when they actually occur (Walker et al., 2004; Zamenopoulos & Alexiou, 2007).

H13: There is no significant relationship between recovery attentiveness that generates pleasurable outcome and the provider’s agility to understand the complainants’ environment.

H14: There is no significant relationship between recovery attentiveness that generates pleasurable outcome and the provider’s anticipatory ability to predict the complainants’

environment.

H15: There is no significant relationship between recovery attentiveness that generates

pleasurable outcome and the provider’s ability to adjust and adapt to the complainants’

environment.

3. Materials and methods

The opposing intellectual traditions adopted in social science inquiries are positivism, anti-positivism, or mixed. However, previous (McColl-Kennedy & Spark, 2003; McCollough et al., 2000; Maxham, 2001; Wirtz & Mattila, 2004) and recent (Kim & Jang, 2016; Liu & Matilla, 2015; Sengupta et al., 2015) service failure and recovery studies adopted scenarios method though Chebat and Slusarczyk (2005) alleged the reaction to scenarios as not real to service failure, and that the manipulated lab situations rarely reflect customers’ actual behaviour. Since this study focuses on the effect of complaint handling of real service failure on resilience, this paper adopts positivist approach and critical incident technique-it followed a realist approach to ontology backed by positivist epistemology, relatively deterministic and nomothetic methods. To test the proposed framework and hypotheses, survey data were collected on negative critical incidents from a population of adult air-travelers and frontline staff in the customer service units of the indigenous airlines currently functional. The adult passengers must have, in the recent times, experienced service mishaps, registered their complaints, and received recovery packs accordingly. We focus on Aero Contractors, Arik Air, and Air-Peace because they are local frontline carriers with no cases of air-mishaps.

We dealt with passengers who had at least three cases of service failures within the last two years, as well as customer service officers with at least three years of experience in handling passengers’ ordeals. Preliminary investigation shows that the FAAN takes charge of service failures for all carriers, and guided by FAAN’s information, there were 55 and 28 customer service staff for Lagos and Abuja airports respectively that fall into this category. All were studied since 83 is quite manageable, aside from the need for greater power of prediction. Lagos and Abuja are the focal points because they have minimal insecurity issues and serve as headquarters for local airlines; and play strong host to national and international airports, huge commercial activities, federal and state ministries, government parastatals, expatriates and all tribes in Nigeria. The paper relied on critical mass theory (see Bingham, 1976) and the assumption that cities with higher socio-economic status are in close proximity and show are more prone to amenity-based values than low socio-economic cities, who often emphasize necessity-based (e.g., innovations designed to correct some specific deficiencies). Further interaction with frontline employees shows that, subject to many factors, these three carriers had a weekly average of twenty (20) cases of service failures that meet our earlier criterion for Lagos and Abuja, which means about two hundred and forty (240) cases monthly.

We developed commonness with the frontline customer service employees and studied service failure cases for two months, which indicates studying 480 passengers through the service employees, who subtly get the disgusted passengers that meet the criterion to fill the questionnaire before leaving. For the passengers, simple random sampling technique was adopted for the selection, given that no complainant was denied the chance to be a part of the process. The respondents were encouraged to fill the questionnaire on firm promise that their responses will form the basis for averting the re-occurrence of their ordeals. The questions in the questionnaire are structured and divided into sections—the first section deals with demographic questions; the second with questions on passengers’ relations with the respective airlines and on critical incidents experienced, and the final section with the critical measurement on a scale to address the conceptual framework and the hypothesized relationships.

3.1. Measures

The constructs in the proposed framework are well-researched, and have validated and revalidated measures and scales in literature. Therefore, 31 scale items for the constructs were developed in the instrument after extensive process of literature review and item selection, as well as reviews by informed persons. The subjects were asked to scale their level of agreement to each observed variable (in terms of how they explain the unobserved variable) on a 5-point Likert-type continuum, from strongly agree (5) through strongly disagree (1). Particularly, the constructs of EM, FAC, EFF and COM were each measured by three items from Karatepe (2006); and for ATT, we collapsed items from Davidow (2003) and Folger and Konovsky (1989) into three scale items. For organizational resilience—AG was measured by five items collapsed from Chu (2015) and Kantur (2015); and then, AA and AD were measured by six and five items respectively from Chu (2015). Content validity and construct validity were assessed; the former explains subjectively the extent to which an instrument is truly a comprehensive measure of the area under study; and the latter deals with the extent to which the items in the scale measure the same construct (Nunnally, 1978).

4. Data analysis and results

4.1. Participants’ profile and normality test

Analysis involved 403 copies of the returned questionnaire that were found good, out of which, 65 percent of the respondents were males, and 35 percent females; 72 percent had education background of at least Bachelor Degree; 30 years and above take about 82 percent of the respondents; people with monthly income of at least N200, 000 form at least 83 percent of the respondents; and finally, 69 percent of the respondents have airline experiences of at least five years. However for each construct, we use a skewness-kurtosis approach to test normality of univariate distribution (Byrne, 2010). Table 1 reports that the tests of statistical skewness and kurtosis have values within the thresholds—all values were less than 3 for skewness, whereas kurtosis values were not more than 8 (Byrne, 2010; Kline, 2011).

Table 1. Test of normality and measurement

Latent variable	Item	Standardized loadings (≥0.50)	Mean	Skewness (≥3.0)	Kurtosis (≥8.0)	AVE (≥0.50)	CA (≥0.70)
Empathy	EMP1	0.908	3.60	1.236	1.200	0.876	0.946
	EMP2	0.969	3.59	1.162	1.002
	EMP3	0.906	3.64	−1.183	0.103
Attentiveness	ATT1	0.885	3.46	1.248	1.408	0.790	0.919
	ATT2	0.912	3.59	1.166	0.122
	ATT3	0.874	3.58	0.210	0.510
Compensation	COM1	0.808	3.36	1.199	1.790	0.699	0.868
	COM2	0.899	3.32	1.182	1.002
	COM3	0.815	3.36	1.217	0.272
Effort	EFF1	0.900	3.38	1.262	1.620	0.841	0.940
	EFF2	0.953	3.43	1.207	1.755
	EFF3	0.897	3.44	0.062	0.034
Facilitation	FAC1	0.863	3.44	1.224	1.342	0.753	0.901
	FAC2	0.900	3.40	1.205	−1.050
	FAC3	0.849	3.45	1.202	1.304
Anticipatory Ability	AA1	0.702	3.28	1.287	1.807	0.822	0.954
	AA2	0.731	3.36	1.207	1.070
	AA3	0.864	3.39	−0.212	−0.372
	AA4	0.889	3.37	1.181	1.412
	AA5	0.827	3.41	1.252	1.521
Agility	AG1	0.856	3.39	1.300	1.004	0.612	0.941
	AG2	0.660	3.49	1.533	1.331
	AG3	0.915	3.43	1.236	1.632
	AG4	0.911	3.44	−0.030	−0.336
	AG5	0.895	3.44	1.241	1.412
	AG6	0.866	3.42	1.271	1.321
Adaptability	AD1	0.687	3.31	1.264	1.641	0.794	0.948
	AD2	0.828	3.42	1.206	1.006
	AD3	0.848	3.38	1.971	1.001
	AD4	0.771	3.43	1.209	1.009
	AD5	0.789	3.38	−0.245	−0.455

N = 403.

Source: SPSS output, 2020.

4.2. Structural equation modeling (SEM) analysis

The conceptual framework and the hypothesized relationships were estimated using SEM; and the population variance-covariance matrix was analysed using AMOS 5.0 (Anderson & Gerbing, 1988; Hair et al., 2011). SEM predominantly suffices when the proposed framework has latent variables with interrelated dependence or a series of causal relationships amongst themselves. Indeed, two-stage SEM was employed—every fundamental construct was subjected to measurement model and then, structural model analysis.

4.2.1. Measurement model

Performing the psychometric evaluation of observed items is critical for SEM (Hair et al., 2011); and with a sample of 403, SEM analysis is good given its sample benchmark of 100 or 150 to 200 and above (Anderson & Gerbing, 1988; Bagozzi, 1984). Table 1 tells about the measurement model and shows that the confirmatory factor analysis model fitness (unidimensionality) was tested in the CFA followed by evaluating the validity and reliability of the multi-item indicators, as well as common method bias (CMB). The model fitness portends evaluation of fit indices using CMIN/DF (≤3.000), NFI(≥0.95), TLI(≥0.95), CFI(≥0.95), GFI(≥0.95) and RMSEA(≤0.06), where CMIN/DF {where CMIN = 638.389; DF = 447; p = 0.000} = 1.428, NFI = 0.995, TLI = 0.990, CFI = 0.997, GFI = 0.952 and RMSEA = 0.069. These suggest that the Chi-square (χ²) is significant and that, the other fit indices met the scientific thresholds and have high adequate goodness of fit with the data gathered; thus, there is no need for re-specifying, re-purifying and improving the model (Byrne, 2010; Hart et al., 2011). These multiple fit indices were used because they provide different information about the model fits-absolute fit, parsimony correction and comparative fit.

Besides, SEM uses these indices because they provide more reliable and conservative evaluations of the solution when used together. The multi-dimensional constructs in the measurement model were further evaluated for the validity and reliability of their scale items. For construct validity, convergent and discriminant were inspected, and the means for the items were good and ranged between medium (2.5–3.4) and high (3.5–5.0) as proposed by Oxford and Burry-Stock (1995) and their AVEs surpassed the 0.5 threshold (Hart et al., 2011). The discriminant validity describes the extent to which a given construct shares more variance with its measures than other latent variables in the model (Fornell & Larcker, 1981; Sanchez & Roldan, 2005); thus, the square roots of constructs’ AVEs in the diagonal exceed inter-construct correlations, affirming discriminant validity (see Table 2). In other words, the indicators loaded onto separate factors in the expected manner and show no oblique but orthogonal relationships among themselves. Construct reliability tested the internal consistency of the scale items using Cronbach’s alpha (α) and composite reliability (CR). CR measures shared variance among a set of observed variables measuring an underlying construct (Fornell & Larcker, 1981)—CR value of at least 0.6 for every factor is generally permissible. Further, the observed items are good, given that the degree of association between the underlying latent factors and each item was statistically significant and met the recommended threshold of 0.7 (Nunnally, 1978).

Table 2. Inter-item correlation matrix

/	Ef1	Ef2	EF3	F1	F2	F3	Em1	Em2	Em3	Co1	Co2	Co3	At1	At2	At3	Aa1	Aa2	Aa3	Aa4	Aa5	Ad1	Ad2	Ad3	Ad4	Ad5	Ag1	Ag2	Ag3	Ag4	Ag5	Ag6
Ef1	1
Ef2	0.864	1
EF3	0.814	0.862	1
F1	0.478	0.534	0.465	1
F2	0.514	0.532	0.471	0.774	1
F3	0.533	0.501	0.493	0.71	0.764	1
Em1	0.534	0.52	0.501	0.424	0.447	0.372	1
Em2	0.569	0.565	0.565	0.438	0.468	0.407	0.867	1
Em3	0.537	0.569	0.557	0.409	0.461	0.446	0.785	0.856	1
Co1	0.416	0.445	0.392	0.388	0.449	0.369	0.458	0.472	0.456	1
Co2	0.38	0.389	0.371	0.441	0.479	0.382	0.338	0.413	0.36	0.726	1
Co3	0.287	0.323	0.34	0.376	0.374	0.324	0.234	0.312	0.31	0.572	0.738	1
At1	0.461	0.472	0.449	0.386	0.421	0.357	0.473	0.526	0.515	0.428	0.395	0.258	1
At2	0.484	0.475	0.433	0.364	0.392	0.424	0.505	0.517	0.509	0.379	0.313	0.226	0.806	1
At3	0.483	0.483	0.461	0.422	0.457	0.398	0.459	0.514	0.489	0.407	0.395	0.299	0.784	0.787	1
Aa1	0.429	0.432	0.424	0.418	0.488	0.429	0.382	0.469	0.469	0.369	0.37	0.239	0.466	0.432	0.457	1
Aa2	0.446	0.457	0.436	0.469	0.491	0.441	0.332	0.388	0.39	0.313	0.319	0.257	0.422	0.419	0.467	0.748	1
Aa3	0.452	0.438	0.454	0.409	0.475	0.424	0.384	0.425	0.421	0.376	0.357	0.264	0.453	0.422	0.447	0.812	0.788	1
Aa4	0.436	0.423	0.429	0.435	0.479	0.416	0.364	0.379	0.349	0.344	0.372	0.303	0.417	0.406	0.492	0.767	0.792	0.883	1
Aa5	0.418	0.436	0.437	0.432	0.439	0.384	0.298	0.361	0.346	0.334	0.417	0.34	0.44	0.395	0.465	0.737	0.752	0.826	0.868	1
Ad1	0.439	0.44	0.43	0.338	0.396	0.344	0.373	0.427	0.42	0.373	0.438	0.307	0.443	0.37	0.382	0.491	0.41	0.398	0.372	0.432	1
Ad2	0.506	0.484	0.502	0.401	0.463	0.42	0.421	0.484	0.458	0.395	0.431	0.315	0.442	0.389	0.429	0.48	0.444	0.447	0.408	0.482	0.833	1
Ad3	0.489	0.481	0.503	0.437	0.44	0.449	0.423	0.442	0.472	0.336	0.426	0.325	0.424	0.404	0.459	0.485	0.476	0.47	0.469	0.488	0.759	0.828	1
Ad4	0.441	0.456	0.486	0.424	0.433	0.398	0.418	0.451	0.439	0.349	0.447	0.331	0.39	0.403	0.421	0.504	0.498	0.501	0.477	0.516	0.694	0.799	0.826	1
Ad5	0.476	0.495	0.469	0.443	0.467	0.392	0.399	0.468	0.451	0.351	0.445	0.396	0.449	0.417	0.47	0.486	0.486	0.48	0.451	0.532	0.724	0.794	0.836	0.807	1
Ag1	0.493	0.512	0.499	0.447	0.475	0.406	0.463	0.494	0.454	0.411	0.424	0.27	0.555	0.44	0.485	0.542	0.495	0.52	0.496	0.534	0.574	0.595	0.542	0.567	0.529	1
Ag2	0.365	0.349	0.342	0.29	0.383	0.342	0.339	0.339	0.352	0.296	0.311	0.206	0.359	0.372	0.359	0.371	0.378	0.324	0.378	0.342	0.394	0.419	0.401	0.409	0.433	0.589	1
Ag3	0.472	0.464	0.483	0.461	0.501	0.446	0.442	0.442	0.41	0.404	0.414	0.267	0.453	0.421	0.491	0.508	0.47	0.497	0.497	0.491	0.515	0.544	0.54	0.528	0.521	0.795	0.592	1
Ag4	0.455	0.468	0.464	0.448	0.445	0.429	0.373	0.408	0.43	0.357	0.386	0.254	0.402	0.402	0.458	0.512	0.466	0.506	0.475	0.491	0.482	0.523	0.537	0.568	0.513	0.767	0.564	0.834	1
Ag5	0.463	0.449	0.428	0.366	0.441	0.444	0.34	0.388	0.393	0.35	0.402	0.31	0.411	0.429	0.46	0.482	0.513	0.443	0.464	0.504	0.527	0.586	0.565	0.523	0.533	0.736	0.578	0.802	0.804	1
Ag6	0.415	0.425	0.369	0.435	0.464	0.369	0.371	0.354	0.367	0.334	0.402	0.338	0.385	0.384	0.428	0.432	0.525	0.421	0.449	0.478	0.494	0.497	0.506	0.508	0.544	0.701	0.542	0.748	0.768	0.838	1

4.2.2. Common method bias (CMB)

To ensure the dataset is free from CMB, Harman’s single-factor test was performed (Harman, 1976; Podsakoff et al., 2003). The test suggests that common method variance suffices when one factor accounts for the criticality of the co-variance in the independent and dependent variables. An exploratory factor analysis involving the scale items of the eight constructs in the framework unveiled that the factors explained 62.4 percent (n = 403) of the variance in the constructs. Actually, no single factor emerged as strong as the first factor was able to, given its 36.2 percent variance, which is less than the threshold of 50 percent (Podsakoff et al., 2003). Therefore, the dataset used here does not have issues relating to CMB.

4.2.3. Structural model

The result on validating the measurement model suggests that structural model has sufficient goodness of fit to the observed data {see χ²,CMIN/DF, NFI, TLI, CFI, GFI and RMSEA}and that the hypotheses are testable via assessing the structural model in a covariance-based approach to model causal relationships amongst the variables (Urbach & Ahlemann, 2010). Scholars (Hair et al., 2011; Sun et al., 2013) posit that the covariance-based and variance-based alternative approaches differ in their approximation to SEM in terms of end-point estimates. We controlled the effects of demographic characteristics, given that they often influence certain kinds of behaviour (Ybarra & Suman, 2008). However, the co-variances and/or correlations of the dimensions of complaint handling and measures of resilience attempt to address the research questions. The regression weight estimates of empathy (EM) and resilience (OR) showed EM and organizational agility (OA) {EM<–>OA} have a direct and significant effect {β = 0.633, p < 0.05}; whereas EM and organizational adaptability (AD) {EM <–>AD}at {β = −0.187, p < 0.000}; and EM and anticipatory agility (AA) {EM <–>AA} at {β = −0.183, p < 0.000}have significant inverse directionality. The significant direct effect of EM<–>OA signifies that a unit increase in EM associates with an increase in OA by 63.3 points; and the significant inverse directionality interprets that a unit increase in EM associates with a decrease in AD and AA by 18.7 and 18.3 points respectively.

For facilitation (FAC), its regression weight and directionality estimates on OR report that FAC<–>OA at β = 0.141, p < 0.05 and FAC<–>AD at β = 0.324, p < 0.000 have direct and significant effect, whereas FAC<–>AA has no effect {β = 0.058}. Implicit is that a unit increase in FAC attracts an increase in OA and AD by 14.1 and 32.4 points respectively. Similarly, compensation (COM) has regression weight and directionality estimates on OR: COM<–>OA has no effect {β = 0.089}; COM<–>AD at {β = −0.026} though with inverse directionality has no effect too; and COM<–>AA at {β = 0.516, p < 0.000} has a direct and significant effect, indicating that a unit increase in COM brings about an increase in AA by 51.6 points. The correlation and co-variance of EFF<–>OR showed EFF<–>OA {β = 0.033} and EFF<–>AA {β = −0.058} have no effect; and {EFF<–>AD} at {β = 0.163, p < 0.000} has direct and significant effect, suggesting that a unit increase in EFF attracts an increase in AD by16.3 points. Finally, the regression weight estimates for ATT to OR report thus: ATT<–>AA has no effect; though ATT<–>OA at {β = 0.110, p < 0.000} and ATT<–>AD at {β = 0.367, p < 0.000} showed direct and significant effects—a unit increase in ATT generates a corresponding increase in OA and AD.

5. Discussion

This paper proposed a conceptual framework and showed the correlations and co-variances amongst the constructs in a bid to enrich the theoretical base of complaint management. Given the scholarly submissions (e.g., Chebat & Slusarczyk, 2005; Diaz et al., 2017; Kim & Jang, 2016; Liu & Matilla, 2015; Sengupta et al., 2015), the study’s findings were heuristic and differ a bit from the basic assumption that complaint handling positively associates with firm’s resilience. Fifteen hypothesized relationships were formulated and theoretically developed, the test items for the latent variables confirmed in the measurement model, and the conceptual framework tested using SEM via AMOS. However, not all the hypothesized relationships were statistically supported (see Table 2 and Figure 2) though some exogenous latent factors had negative significant standard estimates (SE). Some of the hypotheses were not statistically critical because of the interconnectivity and overlapped operationalization of measures of the endogenous latent factors—AA, AD and OA; thus, the effect of one measure may have been captured by the others. Besides, the nature of airline operations is such that it is very difficult to dissipate the activities of the three measures of OR. The competition amongst carriers and the need to remain strategic and distinctive suggest prompt response and adjustment to customer dynamics, which in the real sense of it cross-cuts amongst all measures of the endogenous factors. Activities of preparedness to deal with surprises (OA) integrate sense and respond (AA), as well as learning to innovate and adjust to suit the environment (AD).

Figure 2. Algorithm

The empirical test for H01—H03 involved the co-variances of EM and the endogenous factors—OA, AA and AD, though the relationships were significant and mixed—while EM<–>OA had direct relationship, EM<–>AA and EM<–>AD had significant inverse relationship. These suggest that EM correlates with OR—an increase in EM attracts a corresponding effect on OA, and a corresponding reduction effect on AA and AD. Subject to established relationship, proficient expressions of empathy and explanation to the disgusts psychologically reduce the provider’s preparedness to sense-and-respond to consumer dynamics and to alter processes to cope with change. The environment is already known because safety ranks first in most people’s choice of airlines in Nigeria; thus, passengers who register generic and unpreventable (e.g., flight delays and/or cancellations owing to weather conditions) issues rarely switch or pursue recovery to a logical end. Further even when the problems are preventable, passengers committed to a provider have lower recovery expectations and thus, believe that clear explanation and deeper accord may settle-out the ordeals (del Rio-Lanzaet al., 2009; Michel et al., 2009) and therefore, no need unilaterally anticipating a change or thinking of how to cope with change. The Maussian theory of gift-giving (Mauss, 1990) suggests that complainants are specialized gifts that help firms to anticipate and cope with tribe’s dynamic interests.

The direct relationship {EM<–>OA} is consistent with, and the inverse relationship {EM<–>AA; EM<–>AD} contrasts, the findings of extant studies (Hamel & Valikangas, 2003; Tapanainen, 2012; Volberda, 1996; Zamenopoulos & Alexiou, 2007), given that EM is regularly twisted to suit the environment. For H04—H06, FAC correlates with OA, AA and AD: FAC<–>OA and FAC<–>AD relationships are direct and critical, and that of FAC<–>AA showed absence of statistical significance. The findings align with those of other scholars (Conboy, 2009; Tapanainen, 2012; Zamenopoulos & Alexiou, 2007) who emphasize tactical and flexible facilitation to accommodate proactive dealing with consumer complexity. In testing H07—H09, COM correlates with OA, AA and AD, and it was found that COM<–>OA and COM<–>AA were both critical at p < 0.000 though COM<–>AD was not statistically significant.

However, these findings lay credence to previous studies (Chu, 2015; Hamel & Valikangas, 2003; Zamenopoulos & Alexiou, 2007), which suggest that the design of compensation must be regularly adjusted and adapted to tap opportunities and to accommodate customer dynamics. EFF is associated with H10—H12 and its co-variance with OA, AA and AD showed that EFF<–>AD and EFF<–>AA were both statistically significant at p < 0.000; and that of EFF<–>OA was not significant. Lewis and McCann (2004) report that committed effort to redress complainants affects repurchase and resilience. Similarly, frontline officers design and continually redefine their efforts to align and adapt with changes in the environment (Chu, 2015; Stocia et al., 2003; Tapanainen, 2012; Walker et al., 2004). Finally, the results for ATT<–>OA at (p < 0.03) and ATT→AD at (p < 0.000) were statistically significant; and that of EFF<–>AA was not. Davidow (2003) found that attentiveness has the largest effect on resilience, customer satisfaction and repurchase intention. Similarly, scholars (Sengupta et al., 2015; Walker et al., 2004; Zamenopoulos & Alexiou, 2007) found that attentiveness mixed with resilience predicts customer dynamics, discounts their effects on operations, and proposes alternative actions to deal with them.

5.1. Theoretical implications

From a theoretical perspective, this study makes the following contributions. First, the study contributes to, and extends, the theoretical and methodological discourse in services marketing and complaint management domains by developing a conceptual framework and empirically testing it on the airlines in a developing country, given the non-generalizability and contradictory theoretical and empirical results of extant studies (Mostafa et al., 2014; Othman et al., 2014; K. Wang et al., 2014; Y. Wang et al., 2011) from different contexts. Second, unlike the several previous airline studies (Wittman, 2014; Yang et al., 2012; Zhao et al., 2014) and studies in other contexts (Davidow, 2003; Del-Río-Lanza et al., 2013; Maxham & Netemeyer, 2002; Othman et al., 2014; K. Wang et al., 2014; Y. Wang et al., 2011) on complaint management that focused on the proxies of post-complaint behaviour, CSR, and customer satisfaction, this paper provides additional relevance to OR by studying its connection with complaint handling in the context of the airlines. Our results indicate that complaint handling strategies are designed to suit consumer dynamics, given their criticality on resilience. To the best of our knowledge, no previous scholarship has actually developed and tested our kind of framework on airlines. Third, surprises emanated from the findings amidst the peculiarity of the airline services. The inverse relationship for EM<–>AA and EM<–>DA explains the generic nature of some airline complaints, which ordinarily associate less with competitive advantage and rarely cause switching behaviour. When same complaints exist here and there, and the same unpreventable environmental dynamics are involved as it is with the airlines, more empathy will attract less anticipatory ability and adaptability because the providers can rarely predict the environment with precision. Further, the relationships for FAC<–>AA, COM<–>AD, EFF<–>OA, and ATT<–>AA were not statistically supported because of the overlaps of the scale items of the measures of OR.

5.2. Practical implications

The statistical results support the crucial roles of the exogenous latent factors in correlation with the measures of endogenous latent factors. EM <–>OA has the strongest co-variance, followed by FAC <–>OA, ATT<–>OA, and COM<–>OA; COM <–>AA was most critical, followed by EFF<–>AA, and FAC<–>AA, while EM<–>AA had negative significant value; and ATT<–>AD was most critical, followed by FAC<–>AD, and EFF<–>AD, while EM<–>AD was critical at negative value. First, reflecting on these exogenous latent factors suggests repositioning interactivity with, and reinstating, the disgusted passengers, especially the aggressive ones. Laukkanen et al. (2009) posit that one-on-one marketing activity is one best way to persuade disgusted customers to register their complaints and have their issues clearly sorted out by trained frontline officers. In other words, frontline officers should be regularly trained to proficiently handle myriads of passengers’ psychological ordeals with the least supervision amidst extreme pressures. Second, since communication is quite critical, providers are encouraged to regularly use more of social media breakthroughs (YouTube, Facebook, Twitter, Whatsapp, etc.) to interact and deal with customer ordeals cost-effectively.

Third, staff suggestion schemes and customer feedback and suggestion systems via 24 hours telephone hotlines, comment cards and questionnaires, are critical references to understand different customer expectations and opinions, and to develop appropriate standards and policies to deal with each group. Fourth, given the position of COM as the result shows, carriers need to adopt a flexible compensation policy that spans utilitarian and not just the symbolic and psychological in order to build image and positive expectancy disconfirmation. Fifth, EFF, ATT and FAC are quite critical and must be seamless and customer-friendly to deal with customer dynamics and attract repeat behaviour and favourable word-of-mouth publicity.

6. Limitations and future research directions

Although this study represents a fruitful attempt to contribute to knowledge and theory, its results should be interpreted with cautions, as they have some limitations that provide opportunities for further studies. First, like many service failure and recovery studies (de-Río-Lanza et al., 2009; Svari & Olsen, 2012; Wen & Chi, 2013), the sample was moment-specific and the design was cross-sectional, which rarely shows how variables change overtime. The study involved population from two cities (Lagos and Abuja), which in turn, could negatively reflect on the generalizability because the identified causal relationships may be context-based or may lose relevance overtime. The focus on three major airlines could mitigate cross-context applicability and necessitate further research billed to test and extend the capabilities of the instruments. Causal inferences are only made with reference to theory and so, the direction of the causality and long-term applicability of findings may be strengthened when we extend the measures or engage in longitudinal studies, involving quantitative and/or qualitative approaches.

Second, quantitative and/or deductive method(s) was adopted to collect data; thus, testing theory. Future studies should include qualitative or inductive approach with critical incident technique (CIT)—building theories. CIT as successfully used by previous recovery studies (Chelminski & Coulter, 2011; Swanson & Hsu, 2011) provides in-depth insight into the travelers’ perceptions about service failure and recovery experience. Third, the sample description showed that the largest segment of the subject were males of at least 30 years, who are well-educated, have middle-level income, and are experienced in air travels. Therefore, this questions the applicability of the study’s findings and provides scholarly opportunities to replicate the study by taking samples from other population segments that have different demographic characteristics. Fourth, many more dimensions of the independent variable and measures of resilience exist in literature; therefore, further studies are suggested to integrate them in order to improve frameworks for handling air passengers’ ordeals.

7. Conclusion

Scholars emphasize a more comprehensive knowledge of customer attraction and retention (Diaz et al., 2017; Kim & Jang, 2016; Liu & Matilla, 2015; Sengupta et al., 2015; Wittman, 2014; Zhao et al., 2014). Similarly in a bid to build competitive strategy through evolving and maintaining worthy relationships with new and incumbent customers, complaint handling turns a major research and decision interest in services marketing. Often the lifetime in user-firm interface is besieged with unforeseen circumstances of negative expectancy disconfirmation, as well as concomitant recovery plans to reposition trust and satisfaction. Apparently, service mishaps are inevitable and detrimental though a proficiently programmed recovery strategy turns them (the mishaps) into opportunities to build competitive advantage and resilience. This study focuses on the airlines and presents findings based on real failure cases of how complaint handling affects provider’s resilience. The study adopted positivist and critical incident technique to collect data from a sample of 403 respondents made up of frontline officers and air-passengers who actually complained to the carriers in order to stress the import of proficient recovery in shaping consumer behaviour amidst service failures. In conclusion empathy, facilitation, communication, and attentiveness significantly correlated with organizational agility; the same is said of compensation, effort, facilitation, empathy and anticipatory ability; and attentiveness, facilitation, effort, and empathy, and adaptability.

EM <–>OA has the strongest co-variance, followed by FAC <–>OA, ATT<–>OA, and COM<–>OA; COM <–>AA was most critical, followed by EFF<–>AA, and FAC<–>AA, while EM<–>AA had negative significant value; and ATT<–>AD was most critical, followed by FAC<–>AD, and EFF<–>AD, while EM<–>AD was critical at negative value. Table 3 shows the order of strength of the relationships with EM →OA (β = 0.633) having the strongest, followed by COM →AA (β = 0.516), EFF →AA (β = 0.504), ATT →AD (β = 0.367), FAC→AD (β = 0.324), EFF →AD (β = 0.163), FAC →OA (β = 0.141), ATT →OA (β = 0.110) and COM →OA (β = 0.089). Similarly, EM →AA (β = −0.183) and EM →AD (β = −0.187) had significant direct relationships, while FAC →AA, EFF →OA, and ATT →AD had no significant relationships. Therefore, the study concludes that more symbolic and psychological than the traditional utilitarian recovery strategies to failure handling—thus, increase in interpersonal interaction rather than tangible package has greater potential to reduce the level of passengers’ disgusts and increase me-too and lifetime value. This affirms that strategic position and resilience are more guaranteed via opportunities to anticipate and respond to passengers’ dynamics, and to psychologically express and re-open the firm’s strategic windows.

Table 3. Standardized regression weights of the model

S/N	Estimate	Hypothesized relationship	SE	CR	P-value/Label	Decision
1	EM <–> OA (Hypothesis 1)	There is no significant relationship between providers’ empathy in times of service mishaps and their agility to embrace the complainants’ environment.	0.633	11.179	0.01 (**)	Not supported
2	EM <–>AA (Hypothesis 2)	There is no significant relationship between providers’ empathy in times of service failures and their anticipatory ability to embrace the complainants’ environment.	−0.183	−5.591	0.000 (***)	Not supported
3	EM <–>AD (Hypothesis 3)	There is no significant relationship between providers’ empathy to disgusted customer and their ability to adjust to the complainants’ environment.	−0.187	−4.951	0.000 (***)	Not supported
4	FAC <–>OA (Hypothesis 4)	There exists no significant relationship between facilitation that proficiently handles customer ordeals and provider’s agility to embrace the complainants’ environment.	0.141	3.407	0.04 (**)	Not supported
5	FAC <–>AA (Hypothesis 5)	There exists no significant relationship between facilitation that proficiently addresses customer ordeals and the provider’s anticipatory ability to embrace the complainants’ environment.	−0.058	−1.639	0.101	Not supported
6	FAC <–>AD (Hypothesis 6)	There exists no significant relationship between facilitation that proficiently deals with customer ordeals and the provider’s ability to adjust to the complainants’ environment.	0.324	6.486	0.000 (***)	Not supported
7	COM<–>OA (Hypothesis 7)	There is no significant relationship between compensation redress in times of service mishaps and the provider’s agility to embrace the complainants’ environment.	0.089	2.398	0.000 (***)	Not supported
8	COM<–>AA (Hypothesis 8)	There is no significant relationship between compensation redress in times of service failures and the provider’s anticipatory ability to embrace the complainants’ environment.	0.516	8.661	0.000 (***)	Not supported
9	COM<–>AD (Hypothesis 9)	There is no significant relationship between compensation redress in times of service failures and the provider’s ability to adjust to the complainants’ environment.	−0.026	−.658	0.511	Supported
10	EFF <–> OA (Hypothesis 10)	There exists no significant relationship between recovery efforts that generate pleasurable outcome and the provider’s agility to understand the complainants’ environment.	0.033	0.915	0.360	Supported
11	EFF <–> AA (Hypothesis 11)	There exists no significant relationship between recovery efforts that generate pleasurable outcome and the provider’s anticipatory ability to predict the complainants’ environment.	0.504	8.002	0.000 (***)	Not supported
12	EFF <–>AD (Hypothesis 12)	There exists no significant relationship between recovery efforts that generate pleasurable outcome and the provider’s ability to adjust and adapt to the complainants’ environment.	0.163	3.989	0.000 (***)	Not supported
13	ATT <–>OA (Hypothesis 13)	There is no significant relationship between recovery attentiveness that generates pleasurable outcome and the provider’s agility to understand the complainants’ environment.	0.110	2.717	0.04 (**)	Not supported
14	ATT <–>AA (Hypothesis 14)	There is no significant relationship between recovery attentiveness that generates pleasurable outcome and the provider’s anticipatory ability to predict the complainants’ environment.	0.000	0.009	0.993	Supported
15	ATT <–>AD (Hypothesis 15)	There is no significant relationship between recovery attentiveness that generates pleasurable outcome and the provider’s ability to adjust and adapt to the complainants’ environment.	0.367	7.132	0.000(***)	Not supported

Source: Amos 5.0 output on research data, 2020.

Note:*p < 0.01; **p < 0.05; ***p < 0.001.

Additional information

Funding

The authors received no direct funding for this research.

Notes on contributors

Hart O. Awa

Hart O. Awa has L-8298-2019 as a Web of Science (WOS) ResearcherID, powered by Publons of Wiley. He is a professor of Marketing, and a research point-man in the Department of Marketing and Faculty of Management Sciences, University of Port Harcourt, Nigeria. Hart writes and reviews for top-indexed and WOS journals, many of which are housed in Emerald, Taylor & Francis, Elsevier, Wiley, and Sage. He is a member of Emerald Literati Network, UK and Informing Science, US, and a Fellow of Word Business Institute (WBI), Australia.

Chigbo A. Nwobu

Chigbo A. Nwobu is trained in Engineering and Business, and currently the General Manager, Federal Airport Authority of Nigeria (FAAN). He is a DBA student of the Department of Marketing University, Port Harcourt, and has shown momentous scholarly interest in complaint management.

Sunny R. Igwe

Sunny R. Igwe is a senior academic and a prolific scholar in the Department of Marketing University, Port Harcourt, Nigeria.