Assessing the Effects of Service Variability on Consumer Confidence and Behavior

Abstract

This research examines the effects of service variability on consumer confidence and behavior across multiple transactions in a service relationship. This article integrates discussions of service relationships and models of service variability. Leveraging a field study, the authors track over 12,000 experiences across 3,084 consumers for a 2-year period and model the impact of variability in these experiences on consumer relationships. The results reveal variability in a service relationship can significantly impact consumer confidence and that the nature of this relationship is nonlinear, revealing that small variations in quality will have strong effects on confidence and that these effects plateau as variability increases. Despite these overall effects, we also demonstrate that the positive benefits of service improvement strategies can offset these effects. Finally, in a second study, the results suggest firms can insulate themselves from the effects of variability by encouraging consumer involvement in relational investments (i.e., loyalty programs) that provide consumers with both interpersonal (i.e., status) and economic resources (i.e., points). Taken together, the results demonstrate that consumers integrate evaluations across transactions when evaluating a service provider, thus focusing on a single transaction, in isolation, may not accurately capture consumers’ perceptions of the service relationship.

Keywords

services marketing service quality service variability confidence trust customer experience management

The production and consumption of services usually encompasses multiple interactions (Hui and Bateson 1991) and unfolds over a series of consumption episodes (Bolton and Lemon 1999). Across each of these interactions, firms must strengthen consumers’ confidence in the firm’s ability to meet their expectations to foster consumer loyalty (Berry, Carbone, and Haeckel 2002; Rust and Zahorik 1993; Zeithaml 2000). Gwinner, Gremler, and Bitner (1998, p. 104) postulated that “confidence in the provider appear[s] to develop over time.” Thus, we can accurately forecast confidence in a provider only if we understand service excellence across a series of transactions. Despite the fact that confidence and service relationships should inherently develop over time, most research treats consumers’ perceptions of service evaluations as static and focuses on a single transaction (Grönroos 1993; Rust and Oliver 1994). This cross-sectional focus is particularly problematic in a service context due to the inherent variability (Bebko 2000; Murray and Schlacter 1990) and risk (Sun, Keh, and Lee 2012) of service delivery.

By not widening their scope beyond a single transaction, service researchers lack a complete understanding of the development of service relationships. That being said, classic investigations into service can provide a strong baseline to build upon as researchers pursue more comprehensive models of service appraisal. Specifically, early, cross-sectional investigations provide strong evidence that reliability is a core component of service quality evaluations (Parasuraman, Zeithaml, and Berry 1985; 1988) and the provision of service excellence can spill over and have substantial impact on both behavioral outcomes and firm performance (Rust and Zahorik 1993; Zeithaml 2000). While these models provide evidence that isolated transactions can have a profound impact on consumer relationships, they rarely account for the fact that consumers continuously update their beliefs over successive experiences (Hogarth and Einhorn 1992). To truly understand how service relationships evolve, researchers must account for a more holistic evaluation of experiences across a range of transactions and examine how successive experiences can build consumer confidence (Gremler, Gwinner, and Bitner 1998).

Given this gap in the literature, several scholars have provided an initial assessment of how evaluations of service (and its variability) over time can impact consumer behavior (see Table 1). Most relevant to this research, Bolton, Lemon, and Bramlett (2006) demonstrate that extreme variations in quality can impact consumer retention via contracts in a business-to-business (B2B) professional services setting. Similarly, Sriram, Chintagunta, and Manchanda (2015) examine consumer subscription and termination rates for video-on-demand services and demonstrate that service quality improvements have greater impacts on retention for consumers who have experienced less variability in their services. These studies have validated the importance of accounting for variability but both stop short of examining the mediating mechanisms that cause variability to spill over and affect confidence. Prior studies do not examine potential nonlinearities in these relationships, nor do they assess the effects of variability alongside current service evaluations.

Table 1.

Review of Empirical Research on Service Variability/Heterogeneity.

Reference	Design	Context	Key Constructs	Major Findings
Parasuraman, Zeithaml, and Berry (1985, 1988, 1994)	Cross-sectional	Variety of transactional and contractual services	Service quality	Development of SERVQUAL Scale. Reliability is a key dimension of service quality, and firms must honor their promises.
Murray and Schlacter (1990)	Cross-sectional	Products with varying degrees of product/service components	Perceived risk and service level	Products with a higher service component are perceived as having more risk and more product variability.
Bolton and Drew (1991)	Longitudinal	Telephone company	Service quality and change in quality	Service changes have a strong influence on perceptions of overall quality, and the effects of service quality changes become more noticeable in the long run, because perceptions of overall quality change slowly.
Boulding et al. (1993)	Cross-sectional	Educational institutions (professors and staff)	Expectations, perceptions, service quality, and reliability	Overall quality perceptions are primarily influenced by the reliability dimension of service quality.
Bloemer, de Ruyter, and Wetzels (1999)	Cross-sectional	Entertainment, fast food, supermarkets, and health care	Service quality and service reliability	Service reliability has a positive influence on word-of-mouth and purchase intentions.
Rust et al. (1999)	Cross-sectional and longitudinal	Camera batteries and coffee shop	Disconfirmation of expected performance, perceived variance	Reducing product quality uncertainty increases future purchase likelihood, and meeting expectations raises preferences, even if quality perceptions are low.
Bolton, Lemon, and Bramlett (2006)	Longitudinal	Support services for high-technology systems	Contract renewal, design quality, experience quality, and price	Extreme outcomes in quality over time have a significant effect on renewal, and a few instances of exceptional quality exert positive effects on the renewal decision.
Sriram, Chintagunta, and Manchanda (2015)	Longitudinal	Video-on-demand service	Termination rates, customer retention, and service quality	Variability should be included when making inferences about the effects of improvement in average quality; firms should focus on quality improvements for customers experiencing low variability, and low levels of variance drive customer retention.
Ding and Keh (2016)	Cross-sectional	Restaurant, hotel, and fitness club	Customization, standardization, perceived risk, and satisfaction	Heterogeneity in service outputs is not always a problem. Service customization is preferred for people with hedonic tendencies and a need for uniqueness; service standardization is preferred by customers with utilitarian goals and low need for uniqueness.

In the present research, we strive to extend these studies and bridge the gap between empirical models of variability and conceptual models of service relationships by investigating how variability in service delivery over multiple transactions can affect consumers’ perceptions of confidence as well as how this confidence affects purchase behaviors, word-of-mouth (WOM) intentions, and repurchase time frames. In doing so, this research makes several contributions to the literature. First and foremost, we demonstrate that variability in a service relationship can significantly impact consumer confidence and that the nature of this relationship is nonlinear, revealing that small variations in quality will have strong effects on confidence and that these effects plateau as variability increases. Moreover, we provide a balanced assessment of variability by controlling for the effects of the most recent evaluations of the service, thus providing a conservative assessment of the relative effects of recent versus accumulated perceptions. Finally, we demonstrate that firms can insulate themselves from the effects of variability by encouraging consumer involvement in relational investments (i.e., loyalty programs [LPs]) that provide consumers with both interpersonal (i.e., status) and economic resources (i.e., points).

Theoretical Background and Hypotheses

Managing service interactions requires consideration of key characteristics that distinguish them from goods: intangibility, inseparability, perishability, and heterogeneity (Parasuraman, Zeithaml, and Berry 1985). A substantial amount of service research has focused on intangibility and inseparability as vital aspects of services (Lovelock and Gummesson 2004; Zeithaml, Parasuraman, and Berry 1985), whereas heterogeneity and perishability have not received as much attention in the marketing literature (Lovelock and Gummesson 2004). In this study, we focus on heterogeneity because services pose quality control challenges and are difficult to standardize and because predictability in service encounters can lead to greater satisfaction (Delgado-Ballester 2004). These dual challenges reflect a loss of opportunities to create consumer value if not handled effectively. Specifically, we examine how service quality varies across multiple transactions and how it affects consumers’ evaluations of the service provider.

Early research on service heterogeneity demonstrated that heterogeneity problems are greater for labor-intensive services (Zeithaml, Parasuraman, and Berry 1985) in which multiple employees encounter consumers, causing variability in service delivery (Langeard et al. 1981). Yet it is also possible that the service delivery performance of a single employee will vary across encounters (Knisely 1979). To improve our understanding of the process by which variability across service encounters affects consumer attitudes and spending, we develop a conceptual model based on the role of consumer expectations and the relationship of those expectations to confidence and perceived risk in service decision making.

Confidence and Risk Reduction

Service relationships are constantly evolving as consumers accumulate experience with a provider. Across all of these transactions, consumers develop pretrial beliefs about a service (Parasuraman, Zeithaml, and Berry 1994). They then establish and constantly update their performance expectations over time (Bolton and Drew 1991; Boulding et al. 1993). As a result, expectations are crucial to the consumer’s evaluation of overall service quality, especially when determining whether a service met expectations, was delightful, or was a failure. Expectations of service quality are compared with current perceptions, and when expectations are not met, it influences the overall rating of service quality and decreases satisfaction (Bolton and Drew 1991; Oliver and Swan 1989). Negative disconfirmation results in a service failure and can engender negative emotions in consumers (Anderson 1973; McColl-Kennedy, Daus, and Sparks 2003; Sivakumar, Li, and Dong 2014). The opposite (i.e., positive disconfirmation) occurs when service performance exceeds expectations and affects positive emotions in consumers, such as joy or surprise (Oliver, Rust, and Varki 1997).

Disconfirmation has traditionally been examined within a single service exchange, where expectations and perceived quality are assessed during one time period (which we discuss as the focal transaction for the consumer, as opposed to historical or future transactions with the firm). Disconfirmation has been shown to consistently drive changes in satisfaction, yet the disconfirmation process is more dynamic in services of a relational nature. Overall perceptual evaluations of a service relationship are driven by assessments of both the focal transaction and earlier transactions with the firm (Danaher and Mattsson 1994), and these perceptions will also affect future purchases from the firm. The mechanism by which these expectations manifest as decision choices is perceived risk or confidence (Howard and Sheth 1969; Mitchell 1999). Prior research has defined perceived risk as a subjective expectation of loss (Stone and Grønhaug 1993), with an emphasis on the word “subjective,” since consumers typically have little information and their behaviors are often motivated by subjective impressions, contrary to assessments which may be objective (Mitchell 1999). The expectation of loss is also forward-looking, indicating that perceived risk relates to potential future outcomes rather than previous events. Howard and Sheth (1969) go further to discuss perceived risk as the inverse of confidence. When consumers are making decisions, high risk or low confidence can reduce evaluations of a firm and purchase intentions (Johnson and Grayson 2005; Park, Lennon and Stoel 2005).

Ultimately, when examining how consumers evaluate service providers over time, focusing on the impact of service variability on confidence is critical. Confidence is forward-looking and related to the choice a consumer may make, is based on knowledge of the firm’s offerings, and is heavily influenced by prior experiences. By taking a cumulative perspective on service evaluation, we can understand the extent to which prior transactions impact quality evaluations, confidence in the firm, and financial outcomes beyond what may be delivered in the focal transaction (Bolton, Lemon, and Bramlett 2006; Oliver and Burke 1999). In the following section, we provide more specific arguments regarding how variability can impact confidence and, ultimately, consumer outcomes.

Hypothesis Development

In the service literature, confidence has been defined as a feeling of reduced anxiety and comfort in knowing what to expect in a service encounter (Hennig-Thurau, Gwinner, and Gremler 2002). Moreover, it has been hailed as a cornerstone of service relationships developing over time (Gwinner, Gremler, and Bitner 1998). The consistent delivery of high-quality goods and services is the key to building confidence (Chiou and Droge 2006). Additionally, if this excellence in service is repeated over time, frequent, quality encounters can enhance confidence in a provider (Thomas and Menon 2007). To truly understand confidence development in service relationships, scholars must therefore examine how evaluations evolve over time. Specifically, Golder, Mitra, and Moorman (2012) determine that the quality evaluation process begins with transactional episodes and builds over time, such that the initial encounter supports the formation of “evaluated aggregated quality” perceptions of the experience. Over time, perceptions transform into global judgments of confidence based on multiple encounters.

While consistent excellence can increase confidence, variability in the quality of service experienced in the past infuses additional risk into future encounters and can create anxiety in consumers. Thus, increases in variability across accumulated experiences with a service provider can erode the confidence of consumers who may no longer expect to be universally satisfied with a transaction. Figure 1 provides an overview of the research model that captures these effects. Formally, we hypothesize:

Hypothesis 1: Cumulative variability in service quality over multiple transactions reduces consumer confidence.

Figure 1.

Conceptual model.

In addition to the linear effects of variability proposed in Hypothesis 1, we also expect there to be a decreasing incremental effect of variability on confidence. In cases of high variability, where the consumer experiences multiple service delights or failures, there is likely to be an initial shock effect for the consumer when their expectations are disconfirmed. However, beyond the initial, linear effect of variability on confidence, past a certain point the consequences of variability should diminish. We expect this in light of research done on curvilinear effects in relational exchanges, such as that of Sivakumar, Li, and Dong (2014) who use prospect theory to analyze how delights and surprises affect perceived service quality. The authors propose a functional form of perceived service quality which includes diminishing utility, where the initial effects of quality changes on consumer evaluations are large but flatten after the initial shock (see their figure 2, p. 48). We propose a hypothesis following their conceptual model, where the effects of variability on confidence will be steep at first (due to the strength of the linear effects) but diminish at higher values. We hypothesize:

Hypothesis 2: The effect of cumulative variability in service quality on confidence is curvilinear in nature to the extent that variability has a decreasing incremental effect on confidence as variability increases.

While cumulative variability can erode confidence in service, not all variability may be considered universally bad for firms. For example, firms pursuing service improvement might increase variability over prior experiences by delivering better service. In these instances, we would expect consumers to experience feelings of delight in the focal encounter, thus yielding positive effects on confidence development (Oliver, Rust, and Varki 1997). In such instances, although consumers likely will not forget about prior variability in the service delivery, those effects may be overshadowed by recent quality improvements. Given the signaling benefits of service quality improvements and recency effects (Ross and Simonson 1991), we expect that recent improvements can improve confidence in parallel to the eroding effects of historic variability in the relationship. Formally, we hypothesize:

Hypothesis 3: A positive change in service quality over the most recent prior transaction will increase confidence in the service provider.

While the service literature has limited evidence of the drivers of confidence over time, the downstream effects of confidence are better established. Early research shows that confidence benefits are the most important relational benefits across a range of different services (Gwinner, Gremler, and Bitner 1998). Subsequent research reveals that confidence can result in increases in repurchase intentions and WOM (Hennig-Thurau, Gwinner, and Gremler 2002). In an effort to extend these prior studies, we also propose that confidence in a provider spills over and causes increases in consumer spending and WOM intentions.

Hypothesis 4A: Confidence has a direct, positive effect on consumer purchases.

Hypothesis 4B: Confidence has a direct, positive effect on WOM intentions.

Finally, we also expect that confidence in the service provider should reduce repurchase time. Repurchase time captures the total time that elapses between transactions with a service provider (Evans 1994). When consumers are faced with low levels of confidence, they have the option of postponing further purchases, effectively diverting risky choices into the future (Hofstede 1980; Roselius 1971). If a consumer were to use this strategy to cope with lower confidence, this would result in an increase in repurchase times with the firm. However, if consumers are confident in the service they will receive, they are more likely to return in the nearer future. Therefore, we expect a positive relationship between confidence and repurchase time.

Hypothesis 4C: Confidence has a direct, negative effect on repurchase time.

On average, limiting variability in quality evaluations should drive confidence in the service provider, in turn triggering positive outcomes for the firm, but the strength of these effects could be contingent on consumer relationships with the firm. LPs are a mechanism firms use with the goal of building relationships with their consumers (Beck, Chapman, and Palmatier 2015; Viswanathan, Sese, and Krafft 2017), and stronger relationships can have a significant effect on perceptions of variability. We examine the effects of LPs on the relationship between variability and purchase decisions and WOM intentions since LPs are a mechanism that can be initiated and implemented as a managerial choice as opposed to other potential moderators such as consumer traits, which, while potentially significant, are not easily controllable by marketing managers.

One mechanism by which LPs help build overall loyalty is their ability to foster close relationships between the consumer and the firm (De Wulf, Odekerken-Schroder, and Iacobucci 2001), including confidence in the reliability of a firm (Campbell et al. 2010). LPs can create stronger commitment to the firm through several mechanisms. First, simple involvement in an LP can provide consumers some additional value from a firm, which may result in an increase in their overall evaluations of the brand (Voorhees et al. 2015) and create a baseline increase in commitment to the provider. Beyond baseline membership effects, firms can also design programs with various features to induce greater commitment to the firm. Specifically, Chaudhuri, Voorhees, and Beck (2019) demonstrate that providing consumers with earning mechanisms and status benefits through tiers can provide firms with additional gains in program performance.

Earning mechanisms are suggested to enhance commitment to firms in multiple ways. First, the use of a points reward system activates desirable habitual behaviors in members, such as purchasing (Henderson, Beck, and Palmatier 2011). Points generated through LPs also create a switching cost for consumers: Consumers who have gained points via the LP system feel locked in because switching to a competitor would mean losing their points (Dick and Basu 1994). Consumers with LP memberships will avoid seeking alternative firms, even in the face of unfavorable situations (e.g., variability), since they will lose these accrued points and status benefits (Dick and Basu 1994; Mimouni-Chaabane and Volle 2010).

Moving beyond earning mechanisms, many LPs provide their consumers with status benefits through the creation of tiers (Drèze and Nunes 2009). Status is key to further enhancing commitment to the firm because socially relevant benefits, like status, can motivate consumers more than economic stimuli alone (Bateson, Nettle, and Roberts 2006). In an LP context, the addition of tiers and status to a program can create additional relational bonds between consumers and firms and entice consumers to deepen their commitment to the firm in pursuit of higher status benefits (Nunes and Drèze 2006).

Taken together, LPs can enhance commitment to a provider, but not all programs offer the same benefits. As consumers experience more benefits linked to earning mechanisms and status, their commitment should increase toward playing a key role in moderating the effects of variability. Simply put, we expect that the various features of an LP can create laddered increases in commitment to a firm, which can partially buffer the negative effects of variability on consumer intentions. These moderating effects may be explained by the core tenets of intimacy theory, which suggests that as individuals become more attached to a provider, they will be more tolerant of negative experiences in the relationship (Perlman and Fehr 1987).

Intimacy theory was originally designed to explain interpersonal relationships and has been extended to service contexts. Specifically, Voorhees et al. (2009) leveraged intimacy theory to explain how consumer commitment could reduce the effects of waiting time in the contexts of banking, personal service, restaurants, and automotive service. Given this theoretical foundation and prior applications of intimacy theory to services, we expect that LP involvement will make consumers more tolerant of service variability, thus reducing its effects on consumer outcomes. This core proposition is consistent with the findings of Bolton, Kannan, and Bramlett (2000), who demonstrate that LP members discount prior negative evaluations of a provider. Integrating these prior investigations, we believe that the moderating effects of LP involvement are contingent on the commitment-enhancing features of the program to the extent that the moderating effects will be strongest for LPs including a number of mechanisms to increase commitment and weakest when fewer commitment-enhancing features are included. Thus, we propose:

Hypothesis 5: The relationship between service quality variability and confidence is moderated by LP participation to the extent that the negative effects of service quality variability are reduced as a program includes more commitment-enhancing features (i.e., general participation, earning mechanisms, and status).

Study 1: Effects of Variability on Confidence and Purchases

Data

The data for Study 1 were collected from a large portrait studio over a 28-month period. The photo industry is high involvement, making it more likely consumers can recall their prior experiences when making judgments. We used a sample frame of consumers who had completed surveys regarding at least three photo sittings during the 28-month data collection period. The three-visit minimum enabled us to establish the means and variance across transactions to operationalize variability. That is, the first two transactions served to calculate the coefficients of variation (COVs; discussed below), while any subsequent visits provided the data for additional variability calculations, measures of confidence, WOM intentions, spending measures, and repurchase times. Ultimately, the final data set (after removing the first two observations, since variability would be absent) featured 5,845 transactions with 3,084 consumers. This resulted in a multilevel data set where 1,422 (46.1%) of the 3,084 consumers had multiple observations after accounting for the three-transaction minimum. The average number of transactions was 3.90, and 1,422 consumers had at least four transactions.

Operationalization of Variables

Service Quality

Service quality has been conceptualized as featuring two core dimensions based on Grönroos’s (1984) work: technical quality and functional quality. Consumers are interested in both the quality of the final good they receive (technical quality) and the way the service was handled (functional quality), which in total captures the elements of service quality. In our data, the consumers provided overall evaluations of technical quality (quality of portraits) and functional quality (service provided by the photographer) on 5-point scales (1 = poor, 5 = excellent). Assessing the overall excellence of the service experienced is consistent with the measurement approach advocated by Brady and Cronin (2001).

Confidence

Confidence was measured on a 5-point scale that asked the respondents, “Please rate how confident we made you feel that you were going to have a satisfying experience with us.” This item is consistent with measures used by Delgado-Ballester, Munuera-Aleman, and Yague-Guillen (2003) that reference “feeling confidence” and measures by Gwinner, Gremler, and Bitner (1998) that discuss “confidence the service will be performed correctly.” The use of a single-item measure for confidence seems appropriate in this field context, as additional items add little information while potentially aggravating the respondents (Drolet and Morrison 2001).

Cumulative Variability

Leveraging the quality scores, we calculated COVs to assess the variability in service across all preceding service transactions. The COV is the standard deviation divided by the mean (Bolton and Shankar 2003). Accounting for mean quality scores in the COV also corrects for differences in average quality levels and supports comparisons across different quality levels (Bolton and Shankar 2003; Tarasi et al. 2011). In addition, the COV offers a “best predictor” for a model such as ours that includes transactions nested within individuals (Roberson, Sturman, and Simons 2007, p. 585). The COVs were calculated across transactions for each consumer, ranging from 3 to 12 transactions per consumer over the 2-year period. We used all experiences, including the focal transaction, to calculate the accumulated variability in quality across previous experiences. We also used the squared COV terms to test our curvilinear hypotheses.

Relative Change in Quality

In addition to these cumulative variability measures, we included a measure of relative change in quality that captured the extent to which a focal transaction offered a level quality different from the previous transaction. Thus, while the cumulative variability score captures variability across all prior transactions, the relative change is simply a contrast score that captures the differences in quality scores from the most recent experience in the panel and the immediately preceding service experience. This allowed us to assess the effects of the most recent variation in quality versus the cumulative assessment.

Dependent Variables

Our research consists of three dependent variables. First, we measured the total number of portraits purchased, which was counted for the focal transaction. Second, we measured the consumer’s intentions to share WOM about the firm. Consumers were asked, “How likely are you to recommend [the firm] to your friends?” WOM intentions were measured on a 1–5 scale, where 5 = would definitely recommend and 1 = definitely not. Third, we calculated the consumer’s repurchase time as the number of days between the focal transaction and the prior purchase from the portrait studio.

Control Variables

To provide a rigorous test of our hypotheses and to rule out alternative explanations, we used multiple control variables in our analysis. We included relationship length, which was measured as the number of days between the first transaction (of the original data set) and the focal transaction. This accounts for any perceptual effects that may have developed as the relationship with the firm has grown. We also controlled for the scope of the portrait session by including the total number of photos made as a covariate when predicting the number of items sold. Table 2 details the correlations of these variables. Full details on measurement are listed in Appendix Table A1.

Table 2.

Study 1: Descriptive Statistics and Correlations.

Variables	Mean	SD	1	2	3	4	5	6	7	8	9	10	11
1. COV_{Technical quality}	0.12	0.14
2. COV_{Functional quality}	0.11	0.12	.51**
3. Relative change_{Technical quality}	0.04	0.87	.06**	.04**
4. Relative change_{Functional quality}	0.02	0.79	.04**	.04**	.51**
5. Confidence	4.36	0.92	−.47**	−.38**	−.49**	−.39**
6. Items purchased	3.89	8.14	.01**	.01**	−.01**	−.02**	.00**
7. Repurchase time	106.25	94.48	.04**	.04**	.03**	.01**	−.06**	.07**
8. Word-of-mouth intentions	4.69	0.71	−.44**	−.34**	−.47**	−.35**	.60**	.00**	−.05**
9. Technical quality	4.50	0.86	−.57**	−.36**	−.76**	−.39**	.73**	.01**	−.06**	.69**
10. Functional quality	4.51	0.78	−.36**	−.54**	−.41**	−.74**	.61**	.01**	−.04**	.56**	.59**
11. Relationship length	275.92	142.54	.06**	.05**	.00**	.01**	−.05**	.14**	.51**	−.02**	−.04**	−.04**
12. Number of transactions	3.95	1.43	.02**	.00**	−.01**	.01**	.05**	.03**	−.19**	.03**	.03**	.02**	.30**

*p < .05. **p < .01 (two-tailed).

Analysis

Given the nested nature of our data (transactions within consumers), we estimated a multilevel path model in Mplus Version 7. This model allows for the simultaneous estimation of all model paths, and controls for unobservable traits inherent to individual consumers over our longitudinal data. Following our analysis using multilevel path modeling, we formally tested the mediation using procedures established by Iacobucci, Saldanha, and Deng (2007). The results of this analysis are in Table 3.

Table 3.

Study 1: Results.

Predictor	Model 1		Model 2		Model 3
Predictor	Estimate	SE	Estimate	SE	Estimate	SE
Dependent variable = Confidence
Intercept	4.89**	0.01	0.42**	0.09	0.20	0.11
Technical quality COV	−2.12**	0.27	−0.58**	0.02	−0.56**	0.19
Functional quality COV	−3.43**	0.25	−0.94**	0.17	−0.75**	0.18
Technical quality COV²	1.98**	0.70	1.20**	0.42	1.14**	0.42
Functional quality COV²	2.29**	0.57	1.21**	0.33	1.17**	0.33
Technical quality (focal transaction)			0.31**	0.02	0.30**	0.03
Functional quality (focal transaction)			0.57**	0.02	0.63**	0.03
Relationship length			0.00	0.00	0.00	0.00
Transaction number			0.02**	0.01	0.02**	0.01
Relative change in technical quality					0.01	0.02
Relative change in functional quality					−0.05**	0.02
Dependent variable = Number of items purchased
Intercept	−0.06	0.28	−0.08	0.30	−0.09	0.30
Confidence	0.15*	0.07	0.16*	0.08	0.16*	0.08
Number of items made	0.45**	0.01	0.45**	0.01	0.45**	0.01
Dependent variable = Word-of-mouth intentions
Intercept	2.75**	0.07	2.74	0.30	2.74**	0.07
Confidence	0.45**	0.02	0.45**	0.02	0.45**	0.02
Dependent variable = Repurchase time
Intercept	132.16**	6.55	132.17**	6.55	132.17**	6.55
Confidence	−5.95**	1.46	−5.95**	1.46	−5.95**	1.46

Note. R ² values for Model 3: Confidence = 59.5%, items purchased = 18.1%, word-of-mouth intentions = 39.7%, and repurchase time = 0.3%. COV = coefficient of variation; SE = standard error.

*p < .05. **p < .01.

Results

Table 3 provides an assessment of three models with increasing complexity to provide readers a better sense of how the variability effects evolve as we account for other drivers. In the interest of parsimony, we focus on the discussion of the results from Model 3 that feature all proposed effects and the effects of control variables. Overall, the variables in Model 3 explained 59.5% of the variance in confidence, 18.1% of the variance in items purchased, 39.7% in WOM intentions, and 0.3% of the variance in repurchase time.

Predicting Confidence

As far as specific effects, we begin by baselining the models and reviewing the main effects of technical and functional quality from the focal service experience. Both technical (β = 0.30, SE = 0.03, p < .01) and functional quality (β = 0.63, SE = 0.03, p < .01) had strong, positive, direct effects on confidence, providing evidence that a recent excellent service experience can contribute strongly to confidence development. Moving on to the first hypothesis, the results provide support for the effects of variability on confidence, as the COVs for both technical quality (COV_Technical) and functional quality (COV_Functional) had significant direct effects on confidence (COV_Technical: β = −0.56, SE = 0.19, p < .01; COV_Functional: β = −0.75, SE = 0.18, p < .01). These results provide support for Hypothesis 1. To check the results of the curvilinear hypotheses, we find that the quadratic terms for COV_Technical (β = 1.14, SE = 0.42, p < .01) and COV_Functional (β = 1.17, SE = 0.33, p < .01) are both significant, supporting Hypothesis 2. Given the sign and significance of these effects, the results suggest that subtle increases in variability are associated with sharp declines in confidence that diminish as the changes increase. This pattern of results is consistent with the effects proposed in Hypothesis 2. Hypothesis 3 proposed that a positive change in service quality over the most recent prior transaction will increase confidence in the service provider. The results revealed that for technical quality this effect was not significant (β = −0.01, SE = 0.02, p > .05), but for functional quality, we found a surprising negative effect (β = −0.05, SE = 0.02, p < .01), and despite the strong positive benefits of service excellence found in the baseline effects of quality, even a relative improvement in quality can somewhat erode confidence. Thus, Hypothesis 3 was not supported.

Predicting the Dependent Variables

Building on these results, we find that confidence had a significant and direct effect on our three financial outcome variables of interest. Confidence directly affected the number of units purchased by the consumer (β = 0.16, SE = 0.08, p < .05), WOM intentions (β = 0.45, SE = 0.02, p < .01), and repurchase time (β = −5.95, SE = 1.46, p < .01), which supports all three parts of Hypothesis 4. These effects reinforce the critical importance of confidence in service relationships.

Assessing Mediation

Given the nature of the proposed relationships, the effects imply that confidence mediates the effects of variability on the outcome variables. To assess this possibility, we conducted post hoc analyses to formally test the mediation of confidence in our model. To do this, we followed procedures outlined by Iacobucci, Saldanha, and Deng (2007). First, we fit a model where the hypothesized effects (the COVs and the curvilinear effects) were simultaneously estimated as both direct and indirect effects on the three dependent variables, via the confidence mediator. Next, we calculated the z value as a measure of relative sizes of the indirect effects to the direct effects. Mediation was supported for the relationships between technical quality and all three dependent variables (Items Sold: z _{COV_Technical} = 1.82, p < .05; Repurchase Time: z _{COV_Technical} = 3.04, p < .01; WOM intentions: z _{COV_Technical} = 7.55, p > .01). For functional quality variability, results were similar and mediation was supported for the relationships between functional quality and all three outcomes (Items Sold: z _{COV_Functional} = 1.85, p < .05; Repurchase Time: z _{COV_Functional} = 3.20, p < .01; WOM intentions: z _{COV_Functional} = 12.24, p < .01).

Robustness Analyses

To provide a comprehensive assessment of variability and to rule out other potential explanations, we conducted robustness analyses for alternative models. To assess whether the effects of variability were contingent on the time lapse between transactions, we included time between visits and its interaction with the COVs. This model did not find a significant effect of time since prior visit, its interaction with technical variability, or its interaction with functional variability on confidence.

Next, we assessed whether the historical trend of service quality may impact confidence to determine whether this has a more pronounced impact than relative change in quality. We estimated the slope of the changes in quality scores over transactions (i.e., regressed quality scores on the transaction number) and saved these estimates to be used as inputs for the model. For these estimates, positive coefficients are indicative of an improving trend and negative coefficients are indicative of a decreasing trend. Both quality variables revealed a small positive trend in the data (0.16 for technical quality and 0.13 for functional quality). After calculating these trend variables, we added them to the base model along with the focal quality variables, linear and quadratic terms for the COVs, and the control variables. The results revealed that neither trend variable had a significant main nor interaction effect (with the COVs) when predicting confidence. These results suggest that cumulative variance, in general, harms confidence and that the negative variability effect is unaffected by the trend associated with this variance.

Overall, the robustness analyses provide additional evidence that our presented model best represents the factors affecting the development of confidence in a service relationship and the trade-offs between variability across encounters and excellence in a focal encounter.

Study 1 Discussion

Taken together, the results demonstrate that confidence is dynamically developed. First and foremost, service excellence in the focal encounter remains a primary driver of confidence, suggesting there is no substitute for great service. However, our results also demonstrate that consumers consider more than just the focal experience when assessing their confidence in a provider. Specifically, Study 1 demonstrates that variability influences confidence in several interesting ways. First, for both technical and functional quality, cumulative variability of any type can erode consumer confidence. This effect is strongest for initial levels of variability before the effect plateaus as variability increases. Moreover, the results for functional quality suggest that even potentially positive variability can have negative effects on consumer confidence. Thus, while great service in a focal experience is still the dominant path to confidence, these effects can be undermined by the negative effects of variability that operate in parallel.

Study 2: Moderating Effects of LP Status

Building on the results of Study 1, we developed Study 2 to assess the extent to which the effect of quality variability on confidence is contingent on a consumer’s relationship with the firm via a LP. In doing so, we were able to replicate core effects from Study 1 and test Hypothesis 5 to examine whether consumer involvement in an LP can buffer a firm against the negative effects of variability. The context of the second study was air travel as it is a context where the both status benefits via preferential treatment and economic benefits through miles accrual (Chan, Yim, and Gong 2019).

Method

Experimental Design

Study 2 uses a 2 (Variability: high/low) × 5 (LP Membership: nonmember/general member/points earned/status earned/points and status earned) between-subjects design. The scenario described an experience in which participants are planning to take a flight for a short trip. They find an itinerary that fits their schedule and are considering booking the reservation. Participants were randomly assigned to one of the two variability conditions. Both conditions emphasized that on average service has been good, although individual trip quality varied by condition. In the low-variability condition, the boarding process and trip were consistently adequate. In the high-variability condition, the processes ranged from perfect to chaotic.

For the “LP membership” manipulation, the control condition simply made no reference to an airline LP. In the “general membership” condition, participants were told that they get minor benefits such as special promotions and a monthly newsletter but did not reference any points accumulation or status benefits. For the “points earned” condition, the participant had accumulated 100,000 frequent flier miles but had not flown enough over a 1-year period to earn any status benefits. In the “status earned” condition, the participant had earned Silver Status, which awards them minor benefits but did not have any points available in their balance. The “points and status earned” condition included the same frequent flier miles mentioned previously along with the Silver Status. Following the manipulations, we surveyed participants on variables relevant to our study as described below. To conclude, we conducted manipulation checks and asked for demographic information.

Data

The data to test the moderating effect of rewards program participation were gathered via Amazon Mechanical Turk (MTurk). Participants were adults living in the United States whom we assigned randomly to the variability (high or low) and reward status (nonmember, general member, points earned, status earned, or both points and status earned) conditions. Ultimately, 383 respondents completed the survey and passed the manipulation checks (as described below), with an average of 38.3 respondents per cell. The mean age of the sample was 37 years and the majority of respondents (54%) were women.

Measures

All variables for Study 2 were measured using a 5-point scale and are available in Appendix Table A2. To measure confidence, we used a multi-item scale adapted from measures established by Briñol, Petty, and Tormala (2004). Purchase intentions items were adapted from Hui et al. (2004), where participants were asked to indicate the likelihood that they would purchase the itinerary in the scenario with the three scale items including “unlikely/likely,” “definitely no/definitely yes,” and “inclined not to/inclined to.” WOM intentions were measured with four items from Brown et al. (2005) and Sweeney et al. 2020), where participants rated how frequently they would perform such actions as “Speak positively about the airline to others” and “Recommend this airline to a close personal friend.” Repurchase time was measured by asking participants how many days they would wait before buying another ticket with the airline.

Results

Manipulation Check

To confirm the effectiveness of our variability manipulation, we employed a manipulation check near the end of the survey. We checked the variability manipulation with two items. Participants indicated how much variability they had experienced in the past with this airline on a 5-point scale, where the items were on a bipolar scale and anchored with “not very much/a lot” and “very little/a great deal.” The results differed significantly across the variability manipulations (M _high = 4.03, SD = 1.04; M _low = 2.42, SD = 1.41; t = 12.80, p < .001), confirming that the variability manipulation worked as intended. Additionally, at the conclusion of the survey, we asked participants about their relationship with the firm (via LP membership and involvement). We provided five choices corresponding to the five conditions and we removed participants who did not correctly identify their relationship with the firm.

Hypothesis Test

In addition to testing the moderating effects of LP membership on the variability-confidence relationship, we also replicated the results of Hypotheses 1, 4A, 4B, and 4C with purchase intentions, WOM intentions, and repurchase time as the dependent variables. It is important to note that to simplify the experimental design, we examined only the impact of technical quality variability in Study 2. To test the hypotheses, we used PROCESS (Hayes 2017) Model 7 with 10,000 bootstrap samples and we reported the bias-corrected confidence intervals (CI). Specifically, we estimated a regression equation with a binary coded dummy variable for variability (0 = low variability and 1 = high variability) serving as the independent variable, four dummy variables for the five-level LP membership condition where the control (no membership condition) was the reference level serving as the moderator, and an interaction between the variability variable and the LP dummies that all predicted the mediator (i.e., confidence), which then predicted the dependent variables (purchase intent, WOM intent, repurchase time). The results are detailed in Table 4.

Table 4.

Study 2: Results.

Predictor	Estimate	Confidence Interval
Dependent variable = Confidence
Intercept	4.01*	[3.75, 4.27]
Variability	−1.83*	[−2.16, −1.49]
Loyalty program (minor benefits)	0.00	[−0.33, 0.34]
Loyalty program (status)	0.28	[−0.19, 0.75]
Loyalty program (points)	0.25	[−0.08, 0.57]
Loyalty program (points and status)	0.33*	[0.00, 0.66]
Variability × Loyalty Program (Minor Benefits)	−0.08	[−0.53, 0.37]
Variability × Loyalty Program (Status)	0.58	[−0.04, 1.20]
Variability × Loyalty Program (Points)	0.30	[−0.14, 0.74]
Variability × Loyalty Program (Points and Status)	0.48*	[0.04, 0.92]
Dependent variable = Purchase intentions
Intercept	1.37*	[1.11, 1.64]
Variability	−0.18*	[−0.33, −0.03]
Confidence	0.74*	[0.68, 0.80]
Dependent variable = Word-of-mouth intentions
Intercept	0.84*	[0.51, 1.16]
Variability	−0.15	[−0.33, 0.03]
Confidence	0.69*	[0.61, 0.76]
Dependent variable = Repurchase time
Intercept	113.00*	[90.20, 135.80]
Variability	5.52	[−7.01, 18.04]
Confidence	−9.93*	[−15.14, −4.72]

Note. Confidence intervals are 95% bias-corrected intervals based on 10,000 bootstrapped samples. Reference condition for the loyalty program (LP) effects was the control condition (i.e., no LP).

*p < .05.

The results demonstrated a direct effect of variability on confidence (β_variability = −1.83, CI [−2.16, −1.49]), providing additional support for Hypothesis 1. Moreover, confidence had a direct effect on purchase intentions (β_confidence = 0.74, CI [0.68, 0.80]), WOM intentions (β_confidence = 0.69, CI [0.61, 0.76]), and repurchase time (β_confidence = −9.93, CI [−15.14, −4.72]), providing further support for Hypothesis 4. An examination of simple mediation tests revealed that confidence significantly mediated the effects of variability on purchase intentions (indirect effect_{ViaConfidence} = −1.20, CI [−1.35, −1.05]), WOM intentions (indirect effect_{ViaConfidence} = −1.11, CI [−1.26, −0.97]), and repurchase time (indirect effect_{ViaConfidence} = 16.10, CI [6.77, 25.82]). These mediation effects show additional support for the mediating role of confidence that was established in Study 1.

Next, we examined the effects of LP membership as a moderator of the relationship between variability and confidence. If a consumer is a general member in an LP, there is no significant moderation of the variability effects on confidence (β_{Var × MinorBenefits} = −0.08, CI [−0.53, 0.37]). If the consumer is a member of an LP and has earned points, the interaction effect on confidence remains nonsignificant (β_{Var × Points} = 0.30, CI [−0.14, 0.74]). Similar results hold when only status is earned (β_{Var × Status} = 0.58, CI [−0.04, 1.20]). However, if the consumer has both points and status, the effect of variability is moderated to the extent that having both points accumulated and status can reduce the effects of variability (β_{Var × Points & Status} = 0.48, CI [0.04, 0.92]). As a result, LP participation could reduce, but not completely eliminate, the negative effects of variability on confidence. Taken together, the results provide partial support for Hypothesis 5, where consumers involved in an LP with points and status do not experience as sharp a decrease in confidence following variability.

Discussion

Confidence that a service provider can provide excellent service develops over time and can become a cornerstone of service relationship development. One challenge in developing this confidence is controlling for the heterogeneity that is inherent in services as transactions accumulate. Despite these two universal drivers of service evaluations, prior service research has failed to account for the effects of variability in a service relationship on confidence and the downstream changes in consumer behavior. By filling this gap, we affirm that consumers update their evaluations as transactions accumulate with a provider. As a result, variability in service quality across multiple transactions can impact consumer confidence in the firm, ultimately affecting purchase behavior and willingness to recommend the firm. Interestingly, while cumulative variability does directly erode confidence, these effects can be counteracted when the focal experience is associated with an increase in functional quality relative to historical performances. Variability is therefore not universally bad, because recent positive changes in performance provide direct benefits to the firm in parallel to some indirect damage due to their contribution to cumulative variability. Additionally, we find that if consumers have both status and points benefits in a LP, the effects of variability will be reduced. Therefore, these findings offer key implications for both managers and scholars.

Managerial Implications

Given the results of this research, service managers should be aware that accumulated experiences impact the relationship but that recent experiences are the strongest drivers of loyalty. As a result, standardized excellence needs to be a priority. At the same time, the variability inherent in services can also be offset through relational investments, like a LP. Our results demonstrate that consumers’ perceptions of confidence are updated over time; efforts to improve consistency in the service (and ultimately confidence) will take time. Thus, service managers must be patient in assessing the outcomes of reduced variability in their operations. Fortunately, our results related to the focal change in quality suggests that firms should pursue improvement initiatives, because while they will introduce some variability, in general, consumers do give the firm credit for improvements in service that counteract the damage done by an increase in variability.

Implementing standardization

Edvardsson, Gustafsson, and Roos (2005) and Hartman and Lindgren (1993) note that variability in services is not necessarily a bad thing; it can even be an opportunity to create consumer value when leveraged by efficient personnel. Our research offers some nuanced extensions to this discussion. First, cumulative variability in both technical and functional quality can negatively impact consumer confidence. However, the quality of the focal transaction still has a positive and significant impact on confidence. These results underscore the benefits of being consistently excellent and confirm prior work in the literature emphasizing how service delights can benefit firms (Oliver, Rust, and Varki 1997).

As firms pursue the goal of low variability, high-quality service, one obvious strategy is standardization, which ensures more consistent delivery across transactions and employees. While such approaches could limit the ability of a firm to truly delight a consumer by radically exceeding their expectations, it will result in consistently higher confidence, which can drive increased purchasing. If firms are concerned about their service becoming too mechanized through standardization, they can find ways for their employees to delight consumers in a manner that is not expected to be recurring, such as perks through LPs or invitations to special events that will increase bonds without inflating routine expectations. Investments like these have the potential for spurring one-time increases in confidence while not anchoring the expectation of a higher level of quality.

Service improvement strategy

Firms facing variability issues can also leverage the insights from this research to calibrate improvement strategies. The nonlinear effect of variability shows that after an initial shock, the detrimental effects of variability taper off. As a result, small shifts in variability can lead to a significant reduction in confidence, but this effect softens as variability increases. As a result, if firms need to invest in changing the service, they would be best served by doing so quickly via a major overhaul rather than phasing in changes gradually. A large one-time change will benefit from increases in focal quality, inflicting a single hit to confidence that is not much more detrimental than a minor variability shift.

Utilizing LPs effectively

Reducing variability should be a priority for managers, but perhaps some variability is inevitable. In these instances, our results suggest that firms could leverage LPs to buffer the potential negative consequences of variability. Specifically, by introducing and effectively structuring a firm’s LP through both economic and status benefits, managers should be able to build stronger relationships with consumers and ultimately decrease the effects of variability. Our results show that LPs providing consumers with both points and status are duly beneficial to firms as they directly increase consumer confidence while reducing the negative effects of variability on confidence. Extending the results of this research, similar buffers might be provided when consumers have stronger bonds to the firm that could be reflected in NPS and loyalty scores too as the underlying mechanisms would be similar to those guiding our LP effects.

Theoretical Implications

Confidence is a key driver of service relationships, but little prior research has empirically validated paths to developing confidence across accumulated service experiences. This research demonstrates that consumers do accumulate, recall, and weigh prior experiences when developing confidence evaluations of a firm. However, there is no stronger driver than the most recent experience. Future research should extend these efforts to determine whether there are other variables with more enduring effects across transactions on consumer confidence and whether there are other ways firms can proactively manage these elements. In addition to providing an initial look at drivers of confidence, we underscore confidence’s importance by demonstrating that it can impact more than just intentions. Increased confidence can cause consumers to purchase more and to purchase more frequently. Thus, we extend classic studies in service relationships (e.g., Gwinner, Gremler, and Bitner 1998) in underscoring the importance of confidence.

This research also provides fresh insights into service variability. We extend work by Bolton, Lemon, and Bramlett (2006) by demonstrating that variability in transactional and relational assessments over time can have significant impacts on consumer purchase behavior. Bolton, Lemon, and Bramlett (2006) demonstrated that critical quality variations affect consumer renewal behavior in a contractual setting; we show that even subtle variability in quality over time in noncontractual service industries can affect confidence and consumer behavior. In turn, our study underscores the need for models of service quality to assess variability in service performance over time explicitly, as well as capture measures of confidence as outcomes of this variability. Without accounting for these effects, existing models of service decision making may overstate the importance of cross-sectional evaluations. As a result, we encourage more research that leverage data across a range of transactions or time periods to better understand the dynamics of service relationships (e.g., Cambra-Fierro et al. 2018).

Limitations

Like all research, these studies are not without limitations. While we tracked consumer transactions over a 28-month period, it is possible that longer time horizons or investigations in other industries could result in different results than the effects found here. Moreover, our variability scores capture summary assessments of technical and functional service quality, allowing us to investigate more granular variations. For example, if there were systematic issues associated with variability rather than seemingly random differences in the exchange, they could affect confidence differently. Moreover, our research examined the role of variability in two higher involvement services (photo service and air travel) and with specific types of LPs. Future research could assess the extent to which the results might be replicated in other contexts and LP structures. Finally, we acknowledge that seasonal factors may affect repurchase times, and this may have influenced our repurchase time outcome. Future work could test whether seasonal variability has noteworthy differences from variability as a whole.

Conclusion

Our results demonstrate that variability in services over time is harmful to service providers, because it not only jeopardizes the confidence held by existing consumers but also reduces purchase behavior and WOM intentions. However, our results also reveal that variability is not universally bad. Specifically, variability due to an improvement in quality for a focal transaction can benefit the firm and counteract the negative effects of cumulative variability. Taken together, our results demonstrate that consumer evaluations of service providers are far more complex than traditional, cross-sectional models might suggest. Future researchers therefore need to account for the entire history of the consumer relationship when developing models of service decision making.

Supplemental Material

Supplemental Material, Effects_of_Service_Variability_-_Executive_Summary - Assessing the Effects of Service Variability on Consumer Confidence and Behavior

Supplemental Material, Effects_of_Service_Variability_-_Executive_Summary for Assessing the Effects of Service Variability on Consumer Confidence and Behavior by Clay M. Voorhees, Jonathan M. Beck, Praneet Randhawa, Kristen Bell DeTienne and Sterling A. Bone in Journal of Service Research

Footnotes

Appendix

Table A2.

Study 2: Survey Measures.

Variable	Operationalization	Scale
Confidence	“Following my decision to book with the airline, I felt” confident certain valid sure	1–5, where 1 = not at all and 5 = extremely
Purchase intentions	“Imagine you needed to book a flight like the one described in the scenario. Using the items below, please indicate the likelihood that you would purchase this flight with this airline:” Unlikely…Likely Definitely no…Definitely yes Inclined not to…Inclined to	1–5 with anchors noted in the operationalization column
Word-of-mouth intentions	“Please rate how frequently you would” recommend the airline to close personal friends speak positively about the airline to others mention to others that you used the airline made sure that others know that you chose this airline	1–5, where 1 = never and 5 = frequently
Repurchase time	“Given your typical schedule and the conditions in the scenario, about how many days would you wait before buying another ticket with the airline?”	Sliding scale
Variability condition	Dummy variable: Base condition = low variability	Dummy variable
Loyalty program condition	Dummy variables: Base condition = nonmember	Dummy variables

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Clay M. Voorhees

Supplemental Material

Supplemental material for this article is available online.

References

Anderson

Rolph E.

(1973), “Consumer Dissatisfaction: The Effect of Disconfirmed Expectancy on Perceived Product Performance,” Journal of Marketing Research, 10 (1), 38–44.

Bateson

Melissa

Nettle

Daniel

Roberts

Gilbert

(2006), “Cues of Being Watched Enhance Cooperation in a Real-World Setting,” Biology Letters, 2 (3), 412–414.

Bebko

Charlene P.

(2000), “Service Intangibility and Its Impact on Consumer Expectations of Service Quality,” Journal of Services Marketing 14 (1), 9–26.

Beck

Joshua T.

Chapman

Kelly

Palmatier

Robert W.

(2015), “Understanding Relationship Marketing and Loyalty Program Effectiveness in Global Markets,” Journal of International Marketing, 23 (3), 1–21.

Berry

Leonard L.

Carbone

Lewis P.

Haeckel

Stephan H.

(2002), “Managing the Total Customer Experience,” MIT Sloan Management Review, 43 (Spring), 85–89.

Bloemer

Josee

de Ruyter

Wetzels

Martin

(1999), “Linking Perceived Service Quality and Service Loyalty: A Multi-Dimensional Perspective,” European Journal of Marketing, 33 (11/12), 1082–1106, doi:10.1108/03090569910292285.

Bolton

Ruth N.

Lemon

Katherine N.

Bramlett

Matthew D.

(2006), “The Effect of Service Experiences over Time on a Supplier’s Retention of Business Customers,” Management Science, 52 (12), 1811–1823, doi:10.1287/mnsc.1060.0594.

Bolton

Ruth N.

Shankar

Venkatesh

(2003), “An Empirically Derived Taxonomy of Retailer Pricing and Promotion Strategies,” Journal of Retailing, 79 (4), 213–224, doi:10.1016/j.jretai.2003.09.005.

Bolton

Ruth N.

Kannan

P. K.

Bramlett

Matthew D.

(2000), “Implications of Loyalty Program Membership and Service Experiences for Customer Retention and Value,” Journal of the Academy of Marketing Science, 28 (1), 95–108.

10.

Bolton

Ruth N.

Lemon

Katherine N.

(1999), “A Dynamic Model of Customers’ Usage of Services: Usage as an Antecedent and Consequence of Satisfaction,” Journal of Marketing Research, 36 (2), 171–186, doi:10.2307/3152091.

11.

Bolton

Ruth N.

Drew

James H.

(1991), “A Multistage Model of Customers’ Assessments of Service Quality and Value,” Journal of Consumer Research, 17 (4), 375–384, doi:10.1086/208564.

12.

Boulding

William

Karla

Ajay

Staelin

Richard

Zeithaml

Valarie A.

(1993), “Dynamic Process Model of Service Quality: From Expectations to Behavioral Intentions,” Journal of Marketing Research, 30 (1), 7–27, doi:10.2307/3172510.t.

13.

Brady

Michael K.

Joseph Cronin

Jr . (2001), “Some New Thoughts on Conceptualizing Perceived Service Quality: A Hierarchical Approach,” Journal of Marketing, 65 (3), 34–49.

14.

Briñol

Pablo

Petty

Richard E.

Tormala

Zakary L.

(2004), “Self-Validation of Cognitive Responses to Advertisements,” Journal of Consumer Research, 30 (4), 559–573, doi:10.1086/380289.

15.

Brown

Tom J.

Barry

Thomas E.

Dacin

Peter A.

Gunst

Richard F.

(2005), “Spreading the Word: Investigating Antecedents of Consumers’ Positive Word-of-Mouth Intentions and Behaviors in a Retailing Context,” Journal of the Academy of Marketing Science 33 (2), 123–138.

16.

Cambra-Fierro

Jesús

Melero-Polo

Iguácel

Javier Sese

Doorn

Jenny van

(2018), “Customer-Firm Interactions and the Path to Profitability: A Chain-of-Effects Model,” Journal of Service Research, 21 (2): 201–218.

17.

Campbell

Lorne

Simpson

Jeffry A.

Boldry

Jennifer G.

Rubin

Harris

(2010), “Trust, Variability in Relationship Evaluations, and Relationship Processes,” Journal of Personality and Social Psychology, 99 (1), 14–31.

18.

Chan

Kimmy Wa

Yim

Chi Kin

Gong

Taeshik

(2019), “An Investigation of Nonbeneficiary Reactions to Discretionary Preferential Treatments,” Journal of Service Research, 22 (4): 371–387.

19.

Chaudhuri

Malika

Voorhees

Clay M.

Beck

Jonathan M.

(2019), “The Effects of Loyalty Program Introduction and Design on Short- and Long-Term Sales and Gross Profits,” Journal of the Academy of Marketing Science, 47 (4), 1–19.

20.

Chiou

Jyh-shen

Droge

Cornelia

(2006), “Service Quality, Trust, Specific Asset Investment, and Expertise: Direct and Indirect Effects in a Satisfaction-Loyalty Framework,” Journal of the Academy of Marketing Science, 34 (4), 613–627.

21.

Danaher

Peter J.

Mattsson

Jan

(1994), “Cumulative Encounter Satisfaction in Hotel Conference Process,” International Journal of Service Industry Management, 5 (4), 69–80, doi:10.1108/09564239410068715.

22.

Delgado-Ballester

Elena

(2004), “Applicability of a Brand Trust Scale across Product Categories: A Multigroup Invariance Analysis,” European Journal of Marketing 38 (5), 573–592.

23.

Delgado-Ballester

Elena

Munuera-Aleman

Jose Luis

Yague-Guillen

Maria Jesus

(2003), “Development and Validation of a Brand Trust Scale,” International Journal of Market Research, 45 (1), 35–56.

24.

Wulf

Kristof

Gaby Odekerken-Schröder

Iacobucci

Dawn

(2001), “Investments in Consumer Relationships: A Cross-Country and Cross-Industry Exploration,” Journal of Marketing, 65 (4), 33–50.

25.

Dick

Alan S.

Basu

Kunal

(1994), “Customer Loyalty: Toward an Integrated Conceptual Framework,” Journal of the Academy of Marketing Science, 22 (2), 99–113.

26.

Ding

Ying

Keh

Hean Tat

(2016), “A Re-Examination of Service Standardization versus Customization from the Consumer’s Perspective,” Journal of Services Marketing, 30 (1), 16–28, doi:10.1108/JSM-02-2015-0088.

27.

Drèze

Xavier

Nunes

Joseph C.

(2009), “Feeling Superior: The Impact of Loyalty Program Structure on Consumers’ Perceptions of Status,” Journal of Consumer Research, 35 (6), 890–905, doi:10.1086/593946.

28.

Drolet

Aimee L.

Morrison

Donald G.

(2001), “Do We Really Need Multiple-item Measures in Service Research?” Journal of Service Research 3 (3), 196–204.

29.

Edvardsson

Gustafsson

Anders

Roos

Inger

(2005), “Service Portraits in Service Research: A Critical Review,” International Journal of Service Industry Management, 16 (1), 107–121, doi:10.1108/09564230510587177.

30.

Evans

Richard H.

(1994), “Customers’ Satisfaction and the Replacement Cycle,” Psychological Reports, 75 (3), 1179–1182.

31.

Golder

Peter N

Mitra

Debanjan

Moorman

Christine

(2012), “What Is Quality? An Integrative Framework of Processes and States,” Journal of Marketing, 76 (4), 1–23, doi:10.1509/jm.09.0416.

32.

Grönroos

Christian

(1993), “Toward a Third Phase in Service Quality Research: Challenges and Future Directions,” Advances in Services Marketing and Management, 2 (1), 49–64.

33.

Grönroos

Christian

(1984), “A Service Quality Model and Its Marketing Implications,” European Journal of Marketing 18 (4): 36–44.

34.

Gwinner

Kevin P.

Gremler

Dwayne D.

Bitner

Mary Jo

(1998), “Relational Benefits in Services Industries: The Customer’s Perspective,” Journal of the Academy of Marketing Science, 26 (2), 101–114.

35.

Hartman

David E.

Lindgren

John H.

(1993), “Consumer Evaluations of Goods and Services: Implications for Services Marketing,” Journal of Services Marketing, 7 (2), 4–15.

36.

Hayes

Andrew F.

(2017), Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. New York: Guilford publications.

37.

Henderson

Conor M.

Beck

Joshua T.

Palmatier

Robert W.

(2011), “Review of the Theoretical Underpinnings of Loyalty Programs,” Journal of Consumer Psychology 21 (3), 256–276.

38.

Hennig-Thurau

Thorsten

Gwinner

Kevin P.

Gremler

Dwayne D.

(2002), “Understanding Relationship Marketing Outcomes: An Integration of Relational Benefits and Relationship Quality,” Journal of Service Research, 4 (3), 230–247.

39.

Hofstede

(1980), “Motivation, Leadership, and Organization: Do American Theories Apply Abroad?” Organizational Dynamics, 9 (1), 42–63.

40.

Hogarth

Robin M.

Einhorn

Hillel J.

(1992), “Order Effects in Belief Updating: The Belief-Adjustment Model,” Cognitive Psychology, 24 (1), 1–55, doi:10.1016/0010-0285(92)90002-J.

41.

Howard

John. A.

Sheth

Jagdish N.

(1969), The Theory of Buyer Behavior (No. 658.834 H6), New York: Wiley.

42.

Hui

Michael K.

Zhao

Xiande

Fan

Xiucheng

Kevin

(2004), “When Does the Service Process Matter? A Test of Two Competing Theories,” Journal of Consumer Research, 31 (2), 465–475.

43.

Hui

Michael K.

Bateson

John E. G.

(1991), “Perceived Control and the Effects of Crowding and Consumer Choice on the Service Experience,” Journal of Consumer Research, 18 (2), 174–184.

44.

Iacobucci

Dawn

Saldanha

Neela

Deng

Xiaoyan

(2007), “A Meditation on Mediation: Evidence That Structural Equations Models Perform Better Than Regressions,” Journal of Consumer Psychology 17 (2), 139–153.

45.

Johnson

Devon

Grayson

Kent

(2005), “Cognitive and Affective Trust in Service Relationships,” Journal of Business Research, 58 (4), 500–507.

46.

Knisely

Gary

(1979), “Financial Services Marketers Must Learn Packaged Goods Selling Tools,” Advertising Age, 50 (190), 58–62.

47.

Langeard

Eric

Bateson

John

Lovelock

Christopher H.

Eiglier

Pierre

(1981), Marketing of Services: New Insights from Consumer and Managers. Cambridge, MA: Marketing Science Institute.

48.

Lovelock

Christopher

Gummesson

Evert

(2004), “Whither Services Marketing?” Journal of Service Research, 7 (1), 20–41, doi:10.1177/1094670504266131.

49.

McColl-Kennedy

Janet R.

Daus

Catherine S.

Sparks

Beverley A.

(2003), “The Role of Gender in Reactions to Service Failure and Recovery,” Journal of Service Research, 6 (1), 66–82.

50.

Mimouni-Chaabane

Aîda

Volle

Pierre

(2010), “Perceived Benefits of Loyalty Programs: Scale Development and Implications for Relational Strategies,” Journal of Business Research 63 (1), 32–37.

51.

Mitchell

V. W.

(1999), “Consumer Perceived Risk: Conceptualisations and Models,” European Journal of Marketing, 33 (1/2), 163–195.

52.

Murray

Keith B.

Schlacter

John L.

(1990), “The Impact of Services versus Goods on Consumers’ Assessment of Perceived Risk and Variability,” Journal of the Academy of Marketing Science, 18 (1), 51–65, doi:10.1007/BF02729762.

53.

Nunes

Joseph C.

Drèze

Xavier

(2006), “The Endowed Progress Effect: How Artificial Advancement Increases Effort,” Journal of Consumer Research, 32 (4), 504–512.

54.

Oliver

Richard L.

Burke

Raymond R.

(1999), “Expectation Processes in Satisfaction Formation: A Field Study,” Journal of Service Research 1 (3), 196–214.

55.

Oliver

Richard L.

Rust

Roland T.

Varki

Sajeev

(1997), “Customer Delight: Foundations, Findings, and Managerial Insight,” Journal of Retailing 73 (3), 311–336.

56.

Oliver

Richard L.

Swan

John E.

(1989), “Equity and Disconfirmation Perceptions as Influences on Merchant and Product Satisfaction,” Journal of Consumer Research 16 (3), 372–383.

57.

Parasuraman

Zeithaml

Valarie A.

Berry

Leonard L.

(1988), “Servqual: A Multiple-Item Scale for Measuring Consumer Perceptions of Service Quality,” Journal of Retailing, 64 (1), 12–40.

58.

Parasuraman

Zeithaml

Valarie A.

Berry

Leonard L.

(1985), “A Conceptual Model of Service Quality and Its Implications for Future Research,” Journal of Marketing 49 (4), 41–50.

59.

Park

Jihye

Lennon

Sharron J.

Stoel

Leslie

(2005), “On-Line Product Presentation: Effects on Mood, Perceived Risk, and Purchase Intention,” Psychology & Marketing, 22 (9), 695–719.

60.

Perlman

Daniel

Fehr

Beverley

(1987), The Development of Intimate Relationships. New York: Sage.

61.

Roberson

Quinetta M.

Sturman

Michael C.

Simons

Tony L.

(2007), “Does the Measure of Dispersion Matter in Multilevel Research? A Comparison of the Relative Performance of Dispersion Indexes,” Organizational Research Methods, 10 (4), 564–588, doi:10.1177/1094428106294746.

62.

Roselius

Ted

(1971), “Consumer Rankings of Risk Reduction Methods,” The Journal of Marketing 35 (1), 56–61.

63.

Ross

William T.

Simonson

Itamar

(1991), “Evaluations of Pairs of Experiences: A Preference for Happy Endings,” Journal of Behavioral Decision Making, 4(4): 273–282.

64.

Rust

Roland T.

Jeffrey Inman

Jia

Jinmina

Zahorik

Anthony

(1999), “What You Don’t Know About Customer-Perceived Quality: The Role of Customer Expectation Distributions,” Marketing Science, 18 (1), 77–92, doi:0732-2399/99/1801/0077$05.00.

65.

Rust

Roland T.

Oliver

Richard L.

(1994), “Service Quality: Insights and Managerial Implications from the Frontier,” in Service Quality: New Directions in Theory and Practice, Rust

R. T.

Oliver

R. L.

, eds. Thousand Oaks, CA: Sage, 1–19.

66.

Rust

Roland T.

Zahorik

Anthony J.

(1993), “Customer Satisfaction, Customer Retention, and Market Share,” Journal of Retailing, 69 (2), 193–215, doi:10.1016/0022-4359(93)90003-2.

67.

Sivakumar

Mei

Dong

Beibei

(2014), “Service Quality: The Impact of Frequency, Timing, Proximity, and Sequence of Failures and Delights,” Journal of Marketing, 78 (1), 41–58.

68.

Sriram

Chintagunta

Pradeep K.

Manchanda

Puneet

(2015), “Service Quality Variability and Termination Behavior,” Management Science, 61 (11), 2739–2759, doi:10.1287/mnsc.2014.2105.

69.

Stone

R. N.

Grønhaug

(1993), “Perceived Risk: Further Considerations for the Marketing Discipline,” European Journal of Marketing, 27 (3), 39–50.

70.

Sun

Jin

Keh

Hean Tat

Lee

Angela Y.

(2012), “The Effect of Attribute Alignability on Service Evaluation: The Moderating Role of Uncertainty,” Journal of Consumer Research 39 (4), 831–847.

71.

Sweeney

Jillian

Payne

Adrian

Frow

Pennie

Liu

Dan

(2020), “Customer Advocacy: A Distinctive Form of Word of Mouth,” Journal of Service Research, 23 (2), 139–155.

72.

Tarasi

Crina O.

Bolton

Ruth N.

Hutt

Michael D.

Walker

Beth A.

(2011), “Balancing Risk and Return in a Customer Portfolio,” Journal of Marketing, 75 (3), 18–26.

73.

Thomas

Manoj

Menon

Geeta

(2007), “When Internal Reference Prices and Price Expectations Diverge: The Role of Confidence,” Journal of Marketing Research 44 (3), 401–409.

74.

Viswanathan

Vijay F.

Sese

Javier

Krafft

Manfred

(2017), “Social Influence in the Adoption of a B2B Loyalty Program: The Role of Elite Status Members,” International Journal of Research in Marketing, 34 (4), 901–918, https://doi.org/10.1016/j.ijresmar.2017.08.003.

75.

Voorhees

Clay M.

White

Ryan C.

McCall

Michael

Randhawa

Praneet

(2015), “Fool’s Gold? Assessing the Impact of the Value of Airline Loyalty Programs on Brand Equity Perceptions and Share of Wallet,” Cornell Hospitality Quarterly 56 (2), 202–212.

76.

Voorhees

Clay M.

Baker

Julie

Bourdeau

Brian L.

Deanne Brocato

Joseph Cronin

Jr . (2009), “It Depends: Moderating the Relationships among Perceived Waiting Time, Anger, and Regret,” Journal of Service Research 12 (2), 138–155.

77.

Zeithaml

Valarie A.

(2000), “Service Quality, Profitability, and the Economic Worth of Customers: What We Know and What We Need to Learn,” Journal of the Academy of Marketing Science, 28 (1), 67–85, doi:10.1177/0092070300281007.

78.

Zeithaml

Valarie A.

Parasuraman

Berry

Leonard L.

(1985), “Problems and Strategies in Services Marketing,” Journal of Marketing, 49 (2), 33–46, doi:10.2307/1251563.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB