The Effect of a Service Provider’s Competitive Market Position on Churn Among Flat-Rate Customers

Abstract

Flat-rate pricing, as opposed to charging customers for actual usage, dominates many service industries (e.g., telecommunications, health clubs, and music streaming), and customers often express a flat-rate bias and choose flat rates even if a pay-per-use tariff would be less expensive for them. However, evidence of the effect of this bias on churn is mixed. The competitive market position of a service provider may represent a relevant contingency factor related to this effect; building on attribution theory, the current study predicts that customers attribute their flat-rate bias differently, depending on service providers’ strategic positioning, which leads to varying churn behavior. A survival analysis of approximately 2 years’ transactional data gathered from 21,490 customers of a premium Internet service provider affirms that a flat-rate bias leads to churn in the premium segment. Two experimental studies show that customers of premium service providers attribute their flat-rate bias more externally and exhibit lower fairness perceptions but increased churn intentions compared to low-cost customers who make internal attributions and who thus have less negative perceptions and lower churn intentions. Therefore, premium service managers must proactively manage customers who exhibit flat-rate biases to prevent their negative reactions. Low-cost providers generally have less need for such action and can benefit from flat rates without risking increased churn, despite higher price sensitivity of their customers.

Keywords

flat-rate pricing flat-rate bias churn competitive market position

Flat-rates tariffs¹ charge fixed fees for a service, regardless of actual usage, and are ubiquitous in many service industries. Especially in the telecommunications industry, flat rates are a dominant pricing scheme for landline and mobile communication, Internet, and cable television services. Previous research shows that customers often exhibit a flat-rate bias (Hobson and Spady 1988) such that they “value flat-rate service over measured service even when the bill that the consumer would receive under the two services…would be the same” (Train 1991, p. 211). According to standard economic theory, customers should seek to maximize their welfare such that they would churn to a less expensive pricing plan if they realize they are overspending with a flat-rate tariff (Hobson and Spady 1988). Yet evidence about the consequences of flat-rate biases is mixed. Some studies find no negative effect on customer loyalty (Della Vigna and Malmendier 2006; Lambrecht and Skiera 2006; Wolk and Skiera 2010), but others indicate that flat-rate customers with the wrong calling plan exhibit higher churn rates than customers who choose the best tariff for them (Iyengar et al. 2011; Joo, Jun, and Kim 2002; Wong 2010a, 2010b). Appendix A contains a review of these contradictory findings. For this study, we focus specifically on customers’ churn behavior, rather than tariff switching, because churn rates exert a direct effect on firms’ revenues and can cost service providers billions of dollars each year (Ascarza, Iyengar, and Schleicher 2016).

We propose that a service firm’s competitive market position may explain the seemingly contradictory consequences of flat-rate biases. In competitive marketplaces, service providers generally adopt either a cost leadership or a premium differentiation position (Porter 1980). This positioning tactic determines the company’s value proposition as well as its pricing and advertising strategies (Kalra and Goodstein 1998). Low-cost providers mainly rely on price-oriented, undifferentiated advertising strategies and attract price-aware or price-sensitive consumers with rather low internal reference prices (Popkowski Leszczyc and Rao 1990). According to Santonen (2007), price-sensitive consumers tend to be less loyal and have low price thresholds (Han, Gupta, and Lehmann 2001). In turn, these customers may be more likely to detect their own flat-rate bias, even for relatively small monetary losses. Empirical evidence along these lines would suggest that customers of low-cost providers should react more strongly to an erroneous tariff choice and be more likely to churn. We concur with this general perceptual result, but we predict an opposite behavioral effect, grounded in attribution theory. That is, because of their higher value proposition and high level of customer service (Min et al. 2016), customers of premium providers may attribute their erroneous tariff choice to the firm such that they exhibit an even higher desire to leave this “at-fault” service provider than do customers of low-cost providers. We thus expect a higher churn rate among customers of premium services. With these predictions, we seek to answer three main research questions:

Research Question 1: Does the competitive market position of a service provider moderate the effect of a flat-rate bias on churn such that customers of premium service providers have higher churn probabilities than low-cost customers?

Research Question 2: Can attributions of the source of the flat-rate bias explain this moderation such that customers of premium service providers attribute their flat-rate bias more externally than low-cost customers do?

Research Question 3: If so, can premium service providers buffer the negative effect by mitigating external attributions?

In a preliminary study, using 22 months’ worth of transactional data related to 21,490 customers of a premium Internet service provider (ISP), we provide an initial illustration of these phenomena in a premium segment: Premium service customers who exhibit a flat-rate bias have higher churn rates than customers without such a bias. Moreover, the higher the monetary losses the premium service customers experience due to their flat-rate bias, the more likely they are to churn (Study 1). Then in two experimental studies, we find that the attribution of erroneous tariff choices depends largely on the competitive position of the service provider. With their higher value propositions, premium service providers are subject to more external attributions for poor tariff choices such that their customers express lower fairness perceptions and higher churn probabilities. In the low-cost sector, customers instead attribute the flat-rate bias more to themselves. The higher monetary losses due to a flat-rate bias among premium service customers even increase churn probability (Study 2). However, by proactively issuing plan recommendations, premium service providers can buffer at least some of the negative effects of these external attributions (Study 3).

This article makes several contributions to flat-rate pricing research and highlights important customer loyalty insights for service practice. First, we extend research into the consequences of a flat-rate bias (Della Vigna and Malmendier 2006; Lambrecht and Skiera 2006; Nunes 2000) by introducing service providers’ competitive market position as a relevant contingency factor. Customers of premium providers churn when they realize their flat-rate bias, depending on the amount of their monetary loss due to this bias. By identifying this tendency, we help explain conflicting results in prior bias literature (see Appendix A). Second, we delineate the underlying process that drives this increased churn rate. Customers in the premium segment attribute their flat-rate bias more externally than do customers of low-cost providers who are more likely to regard themselves as responsible. Third, by demonstrating that stronger external attributions foster churn among premium service customers, we contribute to research on pricing biases in general, which tends to neglect customers’ actual attribution behavior.

Our results can also help service managers decide how to deal with customers who exhibit flat-rate biases. Premium service providers should manage them proactively. By informing customers about their underlying flat-rate bias, premium service providers can mitigate external attributions and increase fairness perceptions. In contrast, low-cost providers can profit from their customers’ underlying flat-rate bias. For managers in the latter segment, a flat-rate bias even may be a favorable behavior that they should stimulate.

Theoretical Background and Hypotheses Development

Flat-Rate Pricing

When choosing a service, customers often consider several payment options, including pay-per-use tariffs, variable pricing schemes with a predefined price per consumption unit (Krämer and Wiewiorra 2012), or flat-rate pricing, which charges a predefined fixed fee, regardless of actual usage (Ascarza, Lambrecht, and Vilcassim 2012). Research on flat-rate pricing shows that customers often choose and overvalue a usage allowance that exceeds their needs, and alternative pricing plans could minimize their costs (Hobson and Spady 1988). This phenomenon is referred to as a flat-rate bias (Train 1991). A few studies also indicate the existence of a so-called pay-per-use bias, in which case customers prefer pay-per-use pricing models although a flat rate would save them money (Krämer and Wiewiorra 2012; Kridel, Lehman, and Weisman 1993; Lambrecht and Skiera 2006; Miravete 2002). Lambrecht and Skiera (2006) highlight though that the flat-rate bias is more important, frequent, and persistent over time than the pay-per-use bias is.

Accordingly, many studies confirm the existence of flat-rate biases in various contexts (Della Vigna and Malmendier 2006; Lambrecht and Skiera 2006; Nunes 2000; Prelec and Loewenstein 1998; Train, McFadden, and Ben-Akiva 1987), revealing shares of customers with flat-rate biases that vary from 6% (Miravete 2002) to 76% (Kridel, Lehman, and Weisman 1993); the monetary loss due to this bias spans a similarly broad range. In one study, gym users would save US$500 annually if they switched from a flat-rate to a pay-per-use tariff (Della Vigna and Malmendier 2006). Uhrich, Schumann, and Wangenheim (2013) also find that consuming a service to attain hedonic gratification leads to a higher flat-rate bias than does using a service that meets a utilitarian consumption goal.

The causes of this phenomenon are three flat-rate bias effects (Lambrecht and Skiera 2006): taxi meter, insurance, and overestimation. A taxi-meter effect arises during consumption if consumers do not want to “hear” the meter ticking as they use the service because such reminders of the pain of paying lower their enjoyment of the consumption. An insurance effect occurs if customers try to anticipate a risk of overusage due to demand variability that can create financial losses. In these cases, customers trade-off the chance to pay less for more “insurance” that they will not pay more than the flat-rate price. Finally, overestimation means customers simply overestimate their expected usage.

Building on economic theory, customers should strive to maximize their financial benefits and choose the least costly tariff once they realize they are overspending on an existing flat-rate tariff. But evidence is actually mixed. Lambrecht and Skiera (2006) find that customers with a flat-rate bias exhibit substantial tariff-switching behavior but do not churn such that they are “paying too much and being happy about it.” Della Vigna and Malmendier (2006) and Wolk and Skiera (2010) similarly find no negative effects of flat-rate biases on customer loyalty. In contrast, other studies show that flat-rate customers with the wrong calling plan exhibit higher churn rates than customers who chose the best tariff for them (Iyengar et al. 2011; Joo, Jun, and Kim 2002; Wong 2010a, 2010b). According to Joo, Jun, and Kim (2002), 40% of 10,000 mobile telecommunications customers subscribe to the wrong calling plan,² and they also show significantly lower retention rates than those with the economically optimal plan. Wong (2010a) obtains similar results with 1,403 postpaid mobile telecommunications customers. (See Appendix A for an overview of these contradictory findings about the consequences of the flat-rate bias.) Because customer loyalty exerts a direct effect on service providers’ profits, we need better insights into these contradictory findings, which might reflect the influences of various contingency factors. We propose that the competitive market position of the service firm is one such contingency factor that might explain differences in the consequences of the flat-rate bias.

Competitive Market Position

In competitive markets, with discerning and heterogeneous customers, a distinct competitive market position is crucial for service providers. Two business-level strategies can help firms establish this distinction: differentiation or cost leadership (Porter 1980). This dichotomy is widely accepted in management and marketing research as well as practice (Day and Nedungadi 1994; Homburg, Workman, and Krohmer 1999; McAlister et al. 2016). Firms’ competitive market positioning encompasses all their activities, including marketing and distribution (Kalafatis, Tsogas, and Blankson (2000), and also defines their value proposition (Homburg, Workman, and Krohmer 1999; McAlister et al. 2016; McKee, Varadarajan, and Pride 1989), which they then communicate to consumers or potential customers in an attempt to shape their perceptions and expectations (Min et al. 2016; Yoo, Donthu, and Lee 2000). A differentiation or premium strategy offers and promotes services with unique qualities, such as high speed, reliability, or unique customer experiences (Choi, Lee, and Chung 2001), which enables the service provider to charge higher prices than the industry average (Dess and Davis 1984). A cost leadership or low-cost strategy instead offers and promotes services at the lowest price in the market or at least the lowest price-to-value ratio (Porter 1980). Consequently, the relationship between value proposition and market prices depends on the positioning tactic adopted by the firm; they are inseparable (Kalra and Goodstein 1998).

The two distinct competitive strategies also determine the company’s target customer group. Low-cost service providers that implement cost leadership strategies (Choi, Lee, and Chung 2001) attract customers who are very price aware or price sensitive and generally have a lower internal reference price for the services and products in this segment (Hill 1988; Murray 1988). Because “higher price sensitivity implies that consumers attach greater importance to discovering lower prices [, they]…exhibit higher search propensity” (Mehta, Rajiv, and Srinivasan 2003, p. 69) and tend to be less loyal (Santonen 2007). Moreover, price-sensitive consumers have low price thresholds (Han, Gupta, and Lehmann 2001; Monroe 1973). According to such empirical findings, low-cost consumers seemingly should express stronger negative reactions to evidence of their own flat-rate bias and churn more. However, by turning to attribution theory (Heider 1958; Kelley 1967), we actually predict an opposite behavioral effect such that customers of premium service providers are more likely to suffer decreased loyalty.

Hypotheses and Conceptual Framework

Impact of competitive market position on customers’ causal attributions

To understand the perceptions and churn behaviors of customers who realize their own overspending, we build on attribution theory, which predicts that consumers’ causal inferences about a situation determine their responses to it (Heider 1958; Kelley 1967). In marketing literature, attribution theory (Weiner 1985, 1986) frequently provides explanations for consumers’ affective and behavioral responses to certain situations (Bitner 1990), particularly in service contexts. For example, this theory has helped predict consumers’ reactions to service failures (Bitner 1990; Folkes 1984; Hess, Ganesan, and Klein 2007; Van Vaerenbergh et al. 2014), satisfaction in service encounters (Pick et al. 2016; Tsiros, Mittal, and Ross 2004), and fairness perceptions following price increases (Haws and Bearden 2006; Vaidyanathan and Aggarwal 2003). Attribution literature generally distinguishes three dimensions: locus of causality (who is responsible?), locus of control (could the responsible party have avoided the cause?), and stability (is the cause temporary?; Bitner 1990; Boshoff and Leong 1998). Folkes (1988) implies that the dominant causal dimension for customer satisfaction is locus of causality, which is strongly related to postpurchase responses (Oliver 2010). For this research, we accordingly focus on the locus of causality and investigate customers’ attributions for their flat-rate bias, depending on the competitive position of the service provider. Thus, we analyze whether flat-rate customers believe the causes are internal (themselves) or external (the firm).

As Vaidyanathan and Aggarwal (2003) explain, the locus of causality refers to whether the cause of a certain action is internal or external to the person. Attribution theory predicts that people attribute positive, successful outcomes to themselves but negative outcomes to external or situational factors (Peterson et al. 1982). Folkes (1984) finds that customers react more unfavorably when the locus of causality for a product failure is external. If the cause appears firm related, customers expect compensation or an apology. An internal (self) attribution for product failure does not trigger such expectations (Folkes 1984).

Based on attribution theory, we predict differences in the attribution and fairness perceptions of flat-rate-biased customers, depending on the firms’ competitive market positioning (Figure 1). A customer who realizes that he or she is paying too much for a flat rate might develop different attributions and assign fault for the wrong tariff choice to the service provider or to themselves, depending on its market positioning. The attribution then might lead to negative feelings toward the firm or to themselves (Wong 2010b).

Figure 1.

Conceptual framework: Flat-rate biased customers’ perceptions and behavioral intentions.

A firm that adopts a premium, differentiation strategy offers personal connections and customized services, advertises its strong customer orientation to justify its higher prices (Bolton, Warlop, and Alba 2003), and establishes relational bonds with customers that involve trust, commitment, social benefits, and special treatment (Hennig-Thurau, Gwinner, and Gremler 2002). In contrast, low-cost providers work to lower the prices charged to their customers, so they actively seek ways to rely on cost-efficient self-service options rather than tailored services for each customer (Porter 1996). The higher value proposition in the premium segment in turn may induce a stronger external attribution among consumers, whereas low-cost providers, which rely on customers’ self-service provision and have a lower value proposition, are unlikely to cause customers to blame them for poor tariff choices. Formally, we predict:

Hypothesis 1: When premium service customers exhibit a flat-rate bias, they attribute the failure more externally than do low-cost service customers.

Hypothesis 2: When low-cost service customers exhibit a flat-rate bias, they attribute the failure more internally than do premium service customers.

Impact of customers’ causal attributions on affective and behavioral reactions

Bolton, Warlop, and Alba (2003) find that although customers are willing to pay higher prices for products at stores with an expensive image, price increases by premium firms are perceived as less fair because customers attribute price differences between premium and low-cost vendors more to the premium firms’ profit structure and less to their cost structure. We adopt Xia, Monroe, and Cox’s (2004, p. 3) definition of perceived fairness, as a “consumer’s assessment and associated emotions of whether the difference (or lack of difference) between a seller’s price and the price of a comparative other party is reasonable, acceptable, or justifiable.” Following empirical evidence provided by Bolton, Warlop, and Alba (2003), we further argue that the flat-rate bias of a premium customer eventually leads to a decrease in fairness perceptions, due to the high-value proposition and positioning of the premium providers. This positioning can trigger a stronger external attribution for the flat-rate bias, implying essentially that the price paid is too high and that the premium provider is responsible. In line with Bitner (1990) and Folkes, Koletsky, and Graham (1987), we expect an attribution–affect–behavior sequence, in which customers’ external attribution of their flat-rate bias triggers a negative fairness perception, which then leads to increased churn intentions. Customers’ external attribution of their flat-rate bias in turn mediates the impact of the firm’s competitive market position on perceived fairness. Thus, we expect:

Hypothesis 3: When premium service customers exhibit a flat-rate bias, they display decreased perceptions of fairness relative to low-cost service customers, mediated by customers’ external attribution.

Hypothesis 4: When premium service customers exhibit a flat-rate bias, they display increased churn intentions relative to low-cost service customers, mediated by customers’ external attribution and perceived fairness.

Moderating role of amount of monetary losses

Behavioral decision researchers argue that consumer decision-making is strongly affected by emotional aspects (Andersson and Engelberg 2006). For example, the insurance effect arises because of the benefits that stem from a feeling of being safe from bill shocks, so customers can enjoy their consumption more because they do not need to worry about varying or unexpectedly high costs (Lambrecht and Skiera 2006). Depending on the consumer’s appraisal of these psychological benefits (taxi meter, insurance, and overestimation effects), they may justify somewhat higher costs of a flat rate. Yet with increasing monetary losses, those benefits likely diminish. Consumers of premium service providers, paying high prices, likely suffer greater monetary losses due to a flat-rate bias, which should increase the risk of churn in this segment. Therefore, we expect:

Hypothesis 5: The amount of monetary losses due to a flat-rate bias moderates the effect of competitive market position on churn intentions such that the effect of a premium positioning on churn intentions is even stronger for high amount of monetary losses compared to lower amount of monetary losses.

Mitigating customers’ causal attributions and perceptions

Recently, some U.S. service providers (e.g., Verizon) have begun actively recommending different pricing plans to consumers who are subject to a suboptimal tariff (Ascarza, Iyengar, and Schleicher 2016). Noting these developments, we also consider whether customers’ attributions and perceptions vary depending on whether they detect their flat-rate bias themselves or hear about it from a proactive service provider. Aggarwal (2004) argues that customers evaluate and judge service firms according to specific relationship norms. For example, customers in a premium segment might anticipate high relational norms and customized offers that fit their needs and represent a high level of customer service. If the firm does not proactively inform such customers about their flat-rate bias, they may perceive a violation of those relationship norms. Therefore, premium customers’ self-detection of their own flat-rate bias should lead to a stronger external attribution than would detection that results from the service provider’s proactive communication, which in turn implies a negative effect on perceived fairness. Relationship bonds tend to be weaker in the low-cost segment though, so we do not expect a similar effect for these customers. Rather, we predict:

Hypothesis 6: If premium service customers detect their flat-rate bias themselves, rather than through proactive recommendations by the service firm, their perceived fairness decreases, and this effect is mediated by the customers’ external attribution.

Empirical Overview

We propose that competitive market positions represent relevant contingency factors for explaining why certain firms suffer from increased churn due to their customers’ flat-rate bias, while others do not suffer such negative consequences (Figure 1). In turn, we test our conceptual framework and the related hypotheses in three studies. With transactional data, Study 1 offers an initial demonstration of the phenomenon of interest in a premium segment. Study 2 examines whether service providers’ competitive market position triggers different attributions among customers, which then might lead to different behavioral intentions. We also examine if the amount of monetary losses due to the flat-rate bias moderates the relationship between the competitive market position and consumers’ churn intentions. Finally, in Study 3, we investigate a boundary condition to determine whether premium service providers’ active plan recommendations can buffer the negative effects of external attributions to some extent. Together, the three studies provide systematic evidence that premium service providers suffer from the higher churn rate of flat-rate biased customers because those customers attribute their wrong tariff choice more strongly to these providers.

Study 1

We seek initial evidence that customers with a flat-rate bias also exhibit higher churn rates than customers without this bias. With a cohort-based approach, using transactional data from a premium German ISP, we investigate the different churn rates between those two customer groups as well as whether the amount of monetary losses due to an incorrect tariff choice decreases their customer loyalty.

Methodology

We track 21,490 customers who signed up with the focal ISP in the first quarter of our observation window, which features consumer-level data on a monthly basis from January 2003 until November 2004. The premium ISP offered four different tariffs: (1) no fixed fee but a price per minute (.0137€/min), (2) a medium fixed fee (12.88€/month) for a monthly allowance of minutes (30 hour) and a low price per minute if the user exceeds the allowance level (.0137€/minutes), (3) a medium fixed fee (8.57€/month) for a monthly volume allowance in megabytes (1,500 MB) and a low price per additional megabyte beyond the allowance level (.0137€/MB), and (4) a high fixed fee (25.81€/month) for unlimited usage. Customers could monitor their usage on the ISP’s connection manager application and thereby detect suboptimal tariff choices. With no minimum contract duration, they could cancel or switch at any time. Further descriptive information is available in Table 1.

Table 1.

Transactional Data.

	Tariff Options
	Overall	Tariff 1	Tariff 2	Tariff 3	Tariff 4
Number of customers	21,490	5,628	3,289	2,077	10,496
Average customer age (years)	42.0 (12.6)	43.5 (13.1)	43.7 (12.3)	42.7 (12.6)	40.2 (12.3)
Average usage volume (GB)	6.1 (16.2)	0.1 (.3)	0.2 (.7)	0.5 (.8)	12.3 (20.1)
Average usage time (hr)	147.2 (222.3)	6.5 (12.6)	18.8 (20.3)	72.0 (136.5)	277.7 (244.4)
Average invoice amount (Euro)	18.4 (14.7)	7.5 (18.5)	17.6 (14.5)	11.7 (18.1)	25.8 (0)
Average observation period (months)	17.9 (6.3)	16.9 (7.2)	17.9 (6.3)	19.4 (5.1)	17.9 (6.3)

Note. Standard deviations are in parentheses; the allocation of customers to the tariffs is based on the first period. N = 21,490.

Although inclusive allowances, strictly speaking, are two-part tariffs rather than flat-rate tariffs, similar to prior research (Heidenreich and Handrich 2010; Lambrecht and Skiera 2006), we merely distinguish pay-per-use from flat-rate tariffs. Tariff 1 does not include a fixed fee, so it represents a pay-per-use tariff, whereas Tariffs 2 and 3 offer customers considerably higher allowance levels, similar to a flat-rate tariff, and Tariff 4 is a pure flat-rate tariff. Similar to previous transactional data studies, we cannot determine whether customers truly exhibit a psychological bias when they choose a pricing plan that does not match their actual consumption pattern. That is, a suboptimal plan implies the presence of a flat-rate bias, but it is not per se a flat-rate bias. Building on previous research, we assume that a customer exhibits a flat-rate bias if she or he chooses a flat-rate tariff with a higher allowance, although her or his usage patterns indicate that a smaller tariff (i.e., pay-per-use tariff) would have resulted in a lower total invoice. Formally, following Lambrecht and Skiera (2006), we use two criteria to identify a flat-rate bias: The customer exhibits an overall, general flat-rate bias if the pay-per-use tariff would have been less expensive (less strict criteria), and a customer always exhibits a flat-rate bias if the pay-per-use tariff would have been cheaper in every single month (stricter criteria). The always criterion is stricter and encompasses the overall criterion.

To quantify the amount of the flat-rate bias, we determine, for every customer c in each month t, the hypothetical invoice amount ${inv}_{c, t}^{PPU}$ if she or he had chosen the pay-per-use tariff. If a customer exhibits a flat-rate bias, the resulting average monetary loss is the average difference between the actual invoices ${inv}_{c, t}^{FR}$ and hypothetical invoices. For customers without a flat-rate bias, this loss is 0. Thus,

f r b_{c} = {\begin{matrix} {avg}_{t} ({inv}_{c, t}^{PPU} - {inv}_{c, t}^{FR}) \\ 0 \end{matrix} .

Table 2 indicates the extent and prevalence of flat-rate bias in our sample: 41.66% (stricter criteria: 23.97%) of the flat-rate customers in our sample exhibit some flat-rate bias, leading to an average loss of €13.51 (strict criteria: €16.26).

Table 2.

Existence and Amount of Flat-Rate Bias.

	Tariff Structures
	Pay Per Use	Flat Rates
Number of customers	5,628	15,862
Average invoice amount per month (Euro)	8.6 (18.4)	23.9 (7.9)
Overall bias criterion
Customers with flat-rate bias	—	41.66%
Average amount of bias (Euro)	—	13.51 (5.93)
Always bias criterion	—
Customers with flat-rate bias	—	23.97%
Average amount of bias (Euro)	—	16.26 (2.76)

Note. Standard deviations are in parentheses; the allocation of customers to the tariffs is based on the first period. N = 21,490.

Empirical Findings

To determine whether a customer under the wrong pricing plan has higher churn probabilities, we compare the churn probabilities of customers with and without flat-rate biases by calculating the respective proportions, P(churn|FRB) and $P (churn | \bar{FRB})$ , for customers who churned in our observation window. As a control condition, we also calculate the likelihood that consumers with and without this bias switch within the provider’s service portfolio, that is, P(switch|FRB) and $P (switch | \bar{FRB})$ . We test for the significance of these differences using a Mann–Whitney–Wilcoxon test. With Lambrecht and Skiera’s (2006) methodology, we determine that customers who exhibit a flat-rate bias overall have a 2.74% higher churn probability than do customers without a flat-rate bias (Z = −3.84, p < .01). This difference is even higher for the stricter criterion, as Table 3 reveals: Customers who always have a flat-rate bias show an 11.22% higher risk of churn than customers without this flat-rate bias (Z = −12.35, p < .01). Customer switching behavior functions as a control condition; we do not find a significant increase in switching behavior among flat-rate-biased customers.

Table 3.

Differences in Churn and Tariff Switching Probabilities.

	Overall Criterion			Stricter, Always Criterion
	No Bias (%)	Flat-Rate bias (%)	Delta (%)	No Bias (%)	Flat-Rate Bias (%)	Delta (%)
Switch	17.36	16.38	−0.98	16.62	17.86	1.24
Churn	34.28	37.02	2.74***	32.67	43.89	11.22***

Note. N = 15,862.

***p < .01.

Next, to determine whether this increase in churn probability depends on the amount of monetary loss due to the flat-rate bias, we perform a survival analysis, which can handle right-censoring effectively (Li 1995) and exploit duration data fully as a continuous parameter such that it differentiates whether a customer cancels the service in the 2nd or 22nd month. Ignoring such information would reduce the precision of the estimates. For the survival model specification, we employ a parametric Weibull hazard function with shape parameter p, in which the average amount of monetary loss due to a flat-rate bias by a customer c is the sole predictor variable:

h (t) = p t^{p - 1} e^{α_{0} + α_{1} f r b_{u}}

where

h(t) = hazard of a consumer at time t,

t = time,

frb_u = amount of monetary loss due to flat-rate bias of a consumer u, and

pt ^p−1 = baseline Weibull hazard function.

This fully parametric survival model should be more efficient and provide more meaningful results than semiparametric models (May and Hosmer 1998). The estimation of the baseline hazard functions also allowed us to infer actual survival times or churn probabilities. The results of the survival analysis (Table 4) offer initial evidence of our proposed phenomenon: Each additional Euro in monetary losses led to a significant decrease in customer duration by −0.89%.

Table 4.

Survival Model and Logistic Regression Results.

Weibull Survival Model
Variable	α	SE	95% CI	AF	χ²	p
Model 1 (Churn)
Bias amount	−0.01	0.00	[−0.013, −0.06]	0.991	18.31	.000
Scale	0.92	0.01	[0.90, 0.95]
Model 2 (Switch)
Bias amount	0.01	0.00	[−0.00, 0.01]	1.001	0.38	.540
Scale	0.61	0.02	[0.58, 0.65]
Cox Proportional Hazard Survival Model
Variable	α	SE	HR 95% CI	HR	χ²	p
Model 1 (Churn)
Bias amount	.02	.00	[1.01, 1.02]	1.02	47.39	.000
Model 2 (Switch)
Bias amount	−.00	.00	[1.00, 1.01]	1.00	0.50	.479
Logistic Regression Model
Variable	α	SE	OR 95% CI	OR	χ²	p
Model 1 (Churn)
Bias amount	.02	0.00	[1.01, 1.02]	1.02	29.41	.000
Model 2 (Switch)
Bias amount	−.00	0.00	[99, 1.01]	1.00	0.25	.619

Note. CI = confidence interval; AF = acceleration factor; HR = hazard ratio; OR = odds ratio; SE = standard error. N = 15,862.

We confirm the adequacy of our model by plotting the empirically observed Kaplan–Maier cumulative hazards against the Cox–Snell residuals (i.e., cumulative hazard function of the regression model; Hosmer, Lemeshow, and May 2008) and by conducting a Grønnesby and Borgan test (Grønnesby and Borgan 1996; May and Hosmer 1998).

In both models, the scale parameter is <1, indicating a generally increasing trend in the baseline churn probability over time. Previous research in the telecommunications industry similarly shows that churn probability increases over time (Jamal and Bucklin 2006; Li 1995). For our study, this increase might result from the sustainability of the flat-rate bias effect: As a customer gains experience, the insurance effect diminishes, and the propensity to leave increases, in line with economic theory.

To test the robustness of these results, we conducted a sensitivity analysis and estimated the effect of the level of the flat-rate bias with a Cox proportional hazard model and logistic regression (Table 4). A statistically significant effect of the amount of monetary loss due to the flat-rate bias existed only for churn probability; the Cox proportional hazard ratio was 1.015 (p < .0001), and the odds ratio of the logistic regression was 1.015 (p < .0001). In contrast, we did not find a significant effect of the amount of monetary losses on tariff switching behavior. These results confirm that each Euro increase in monetary loss due to a flat-rate bias increases churn probability by approximately 1.5%, in line with our main model.

To conclude, the results also resonate with existing research into the existence of a flat-rate bias and its prevalence (Della Vigna and Malmendier 2006; Kridel, Lehman, and Weisman 1993; Mitchell and Vogelsang 1991; Nunes 2000), in that 42% of the flat-rate customers in our sample pay too much. With regard to the consequences of a flat-rate bias though, our results point in exactly the opposite direction as that suggested by Lambrecht and Skiera (2006). Despite a similar degree of customer reactivity (switching and churn), flat-rate-biased customers in our data reveal a significantly higher churn rate, in preliminary support for our proposed phenomenon in the premium segment. We therefore turn to analyzing the impacts of the firm’s competitive positioning on churn among flat-rate-biased customers, in two experimental scenario studies.

Study 2

Study 1 provides initial evidence of the research phenomena in the premium segment; with experimental studies, we aim to provide a deeper understanding of the underlying psychological mechanisms and test our conceptual framework. Specifically, we investigate how customers with a flat-rate bias attribute that bias, depending on the competitive market position of the service provider. Moreover, we analyze whether this competitive market position triggers different fairness perceptions and churn intentions as well as whether an increase in monetary losses due to the flat-rate bias increases customers’ churn intentions. That is, we test Hypotheses 1–5.

Methodology

Design and Participants

The experiment was a 2 (value proposition: premium/low) × 2 (price positioning: high/low market price) × 2 (level of monetary loss: 20%/40%) between-subject scenario design. We identified mobile phone services as suitable for our experiment; according to telecommunications experts we interviewed, flat-rate tariffs are the most popular pricing plans, and most providers in this industry take clear strategic market positions and proactively communicate their value propositions and prices to consumers. The scenario experiment was conducted online in 2017 by a professional marketing research agency with a sample representative of German mobile phone service customers. The 771 participants had an average age of 40 years, and 49% of them were women.

To manipulate the value proposition and price positioning, we assigned participants randomly to four experimental treatment groups. In each group, participants read a description of a service provider and flat-rate tariff and had to imagine that the displayed flat rate was one that they had chosen previously. We held all information constant except for the two manipulated factors (value proposition and price positioning). According to Scott, Mende, and Bolton (2013), this manipulation is subtle, in that we varied only six words in a scenario description that contained more than 220 words (see the Appendix B for the complete scenario description). The premium value proposition was characterized by its high quality, excellent network coverage, and high level of customer service. The provider with the low-value proposition instead offered average quality and network coverage, a high level of self-service, and a fee-based hotline. The price positioning was manipulated by indicating two different market prices: 39.99€ in the high-price and 14.99€ in the low-price conditions. As manipulation checks, participants indicated the providers’ value proposition according to a 3-item measure from Dabholkar, Shepherd, and Thorpe (2000) and their price positioning on a 3-item measure from Yoo, Donthu, and Lee (2000), both with 7-point scales.

Next, participants read a short statement which indicated they overspent for their current tariff because their consumption patterns failed to match their current allowances. All information was held constant except for the level of flat-rate bias such that we manipulated the monetary amount the customer overspent each month as either 20% and 40%. The different market prices already had been listed, so we simply indicated the level of overspending as a percentage to account for the different costs of the flat-rate tariffs. The manipulation check used a single-choice question related to the different levels of monetary losses. Finally, participants completed an online questionnaire.

Measures and Validity Evaluations

We measured internal and external attribution with two scales adapted from Vaidyanathan and Aggarwal (2003) ³ and Ping (1995), which enabled us to assess churn intentions. For the measures of perceived fairness, we turned to Campbell (2007). In addition, we assessed perceived competence using scales from Fuchs, Prandelli, and Schreier (2010). All the scales relied on multi-item, 7-point scales. To translate the original measurement scales into German, we applied Brislin (1970) back-translation method.

We assessed the convergent and discriminant validity of all the measures using SPSS 24.0 and AMOS 24.0. All measurements exhibited convergent validity, according to the factor loadings, Cronbach’s α, composite reliability, and average variance extracted (AVE), all of which exceeded their well-established limits. To assess the reliability of the 2-item scales, we calculated the Spearman–Brown coefficient (Eisinga, Grotenhuis, and Pelzer (2013). Although Cronbach’s α tends to underestimate the reliability of 2-item scales, our analysis shows that the Cronbach’s α and Spearman–Brown coefficient values are almost the same for internal attribution (.84) and external attribution (.91). Thus, we can assume τ equivalence (Eisinga, Grotenhuis, and Pelzer 2013). As a test of discriminant validity, we used Fornell and Larcker’s (1981) criterion. The square root of the AVE for each 3-item measure was greater than all construct correlations, confirming discriminant validity. Table C1 in Appendix C contains the scales, their origins, and the reliability and validity checks.

Empirical Findings

Manipulation check

Using a multivariate analysis of covariance (MANCOVA), we determined that participants’ perceptions of the value proposition and price positioning differed significantly across the conditions. We find that the price manipulation has a significant main effect on perceived price positioning, F(1, 770) = 265.81, p < .00, M _{High Price} = 4.56, SD _{High Price} = .64; M _{Low Price} = 3.67, SD _{Low Price} =.88, and the value proposition manipulation has a significant main effect on perceived value positioning, F(1, 770) = 228.40, p < .00, M _Premium = 4.90, SD _Premium = 1.50; M _Low = 3.81, SD _Low = 1.50, whereas we did not find any significant interaction effects. The third manipulated factor—namely, level of flat-rate bias—did not exert a significant main effect, neither on perceived price positioning, F(1, 770) = 2.22, p > .10, nor on perceived value positioning, F(1, 770) = 2.15, p > .10. However, 29 participants were screened out after wrongly indicating the amount of monetary losses due to their overspending, which provided an acceptable quality failure rate of <4%.

Preliminary analysis

In a first step, we validate the effects of our manipulations on churn intentions using an analysis of covariance (ANCOVA). The analysis of consumers’ behavioral intention reveals that firms’ value propositions, F(1, 770) = 31.02, p < .00, as well as their price positioning, F(1, 770) = 4.00, p < .05, have significant main effects, but we do not find an effect for the level of flat-rate bias, F(1, 770) = .01, p > .10). Also, no indirect effects were significant. Therefore, as intended, both manipulated factors value proposition and price positioning significantly affect consumers’ behavioral intentions.

We measured the provider’s competitive market position (1 = definitely low cost to 7 = definitely premium) right after the manipulation. An ANCOVA of this competitive market position index reveals significant main effects of the value proposition, F(1, 770) = 70.28, p < .00, and price positioning, F(1, 770) = 144.61, p < .00. A higher value proposition evokes perceptions of a premium provider, more so than a low-value proposition (M _Premium = 4.31, SD _Premium = 1.22; M _Low = 3.60, SD _Low = 1.33). In line with this finding, high market prices are more closely associated with premium service providers, relative to low market prices (M _{High Price} = 4.45, SD _{High Price} = 1.20; M _{Low Price} = 3.42, SD _{Low Price} = 1.23). Thus, both the value proposition and the market price determine perceptions of firms’ competitive market positions.

Moreover, our data clearly show that when the value proposition and price positioning fit (premium/high or low/low), customers can determine the firms’ competitive market position more easily. The combination of a premium value proposition and a high market price is a clear indicator of a premium competitive market position (M = 4.78, SD = 1.07); a rather basic value proposition together with low market prices clearly indicates a low-cost provider (M = 3.04, SD = 1.21). However, the participants struggled to assign the contrary combinations (premium/low or low/high) to any certain strategic positioning (M = 3.95, SD = 1.20). These results are in line with our theoretical conceptualization in which we argue that the value proposition and price positioning are inseparable. To test our hypotheses, we therefore focus on the two experimental groups with matched value propositions and price positioning (N = 391); the other two treatment groups do not allow for a clear distinction between market positions. In line with Porter (1980), we refer to these two focal groups as premium and low-cost positioning (coded as 1 = premium and 0 = low cost).

Internal and external attributions

A MANCOVA of attribution reveals that competitive market position has a significant main effect on external attribution, F(1, 390) = 10.83, p < .05, and a marginal significant effect on internal attribution, F(1, 390) = 3.06, p < .10. Customers of premium providers who exhibit a flat-rate bias attribute their tariff choice more to the provider (M = 3.43, SD = 1.79) than do customers of low-cost providers (M = 2.87, SD = 1.71), in support of Hypothesis 1. In parallel, customers of low-cost providers who show a flat-rate bias attribute their choice to themselves (M = 5.23, SD = 1.56) more than do customers of premium providers (M = 4.97, SD = 1.60) in support of Hypothesis 2. In these tests, we also control for age, gender, income, and perceived competence.

Mediation analysis

To test our Hypothesis 3, we employ a multicategorical mediation analysis as proposed by Hayes (2013), using the PROCESS macro (Model 4). This approach allows for several mediators due to its use of bootstrapping (Preacher and Hayes 2008). Our models rely on the service provider’s competitive market position as the independent variable, perceived fairness as the dependent variable, and internal and external locus of causality as mediating variables. Again, we control for respondents’ age, gender, monthly income, and perceived competence. As we show in Table 5, with a bootstrap confidence interval (CI) of 95% and 5,000 resamples, we find that external attributions negatively mediate the relationship between a service provider’s strategic positioning in the market and perceived fairness (effect = −.09, boot SE = .06, lower level CI [LLCI] = −.23, upper level CI [ULCI] = −.01), in support of Hypothesis 3. That is, flat-rate biased customers of premium service vendors display significantly lower fairness perceptions due to their stronger external attributions.

Table 5.

Results of the Multiple Mediation Analysis.

	Multiple Mediation (Hayes 2013)
	Coeff	SE	t	p
Model: Outcome External Attribution
Competitive market position	.59	.18	3.29	.001 (H₁)
Age	−.14	.18	−.77	.440
Gender	−.01	.01	−1.41	.158
Income	−.06	.05	−1.16	.247
Perceived competence	−.17	.06	−2.74	.008
Model: Outcome Internal Attribution
Competitive market position	−.28	.16	−1.75	.081 (H₂)
Age	.22	.16	1.35	.984
Gender	.01	.01	.85	.179
Income	.05	.05	1.01	.312
Perceived competence	.21	.06	3.73	.000
Model: Outcome Perceived Fairness
External attribution	−.15	.06	−2.50	.013
Internal attribution	.15	.06	2.16	.031
Competitive market position	−.53	.15	−3.53	.000
Age	−.22	.15	−1.48	.140
Gender	−.01	.01	−1.83	.069
Income	−.01	.04	−.20	.841
Perceived competence	−.15	.05	−2.93	.004
Total effect	−.66	.15	−4.27	.000
R ²	.15
F(7, 383)	9.46
p	.000
	Indirect Effects Through Bootstrapping
			95% CI
	Data	SE	LL	UL
Total	−.13	.06	−.26	−.03
External attribution	−.09	.06	−.23	−.01 (H₃)
Internal attribution	−.04	.04	−.15	.00

Note. N = 391.

To further evaluate the robustness of our proposed sequences (competitive market position—attribution—perceived fairness), we tested the alternative sequences (competitive market position—perceived fairness—attribution) by applying the same method. The analysis revealed a stronger direct and total effect under our proposed mediation order.

Using Hayes’s (2013) Model 6 and conducting a serial mediation analysis, we find that the relationship between a service provider’s competitive market position and churn intentions is positively mediated by external attributions and perceived fairness (effect = .02, boot SE = .02, LLCI = .00, ULCI = .06), in support of Hypothesis 4. Customers of premium providers who have a flat-rate bias are at significantly greater risk of churning to competitors due to their increased external attributions and decreased fairness perceptions.

Moderating role of the level of flat-rate bias

We test whether an increase in monetary losses due to a flat-rate bias increases churn among premium customers using an ANCOVA of churn intentions. In line with our previous analysis, we control for age, gender, monthly income, and perceived competence. The analysis reveals a significant main effect of competitive market positioning on churn, F(1, 390) = 6.66, p < .01, and a marginal significant interaction effect between competitive market position and level of flat-rate bias, F(1, 390) = 3.05, p < .10, but we do not find a significant main effect of the level of flat-rate bias (F < 1). That is, premium service providers suffer increased churn intentions among their flat-rate-biased customers when the monetary losses increase (M _Low = 5.29, SD _Low = 1.23; M _High = 5.58, SD _High = 1.27), but low-cost providers do not confront such increased churn risk (M _Low = 5.18, SD _Low = 1.50; M _High = 5.00, SD _High = 1.41). This finding supports Hypothesis 5 and confirms our findings from Study 1.

In conclusion, this experimental study highlights that the consumers’ reactions to their tariff choice bias depend largely on the competitive position of the service provider. Premium service providers offer and communicate higher value propositions and thus are subject to more external attributions for poor tariff choices such that customers of premium providers express decreased fairness perceptions and higher churn probabilities. Among premium service customers, higher monetary losses due to a flat-rate bias further increase churn intentions. In contrast, in the low-cost sector, customers attribute the flat-rate bias more to themselves, and hence there is also no interaction effect with level of flat-rate bias.

Study 3

In line with attribution theory, premium service customers with a flat-rate bias exhibit higher churn intentions than customers of low-cost providers. In Study 3, we thus investigate premium service providers’ active plan recommendations, as a factor that might influence customers’ causal attributions and perceptions, such that we test Hypothesis 6 in this experimental study. We also rule out potential validity concerns regarding participants’ self-selection because the respondents completed the study questionnaire in reference to their actual mobile phone service provider (cf. Study 2, which was based on a random allocation of participants to treatment groups).

Methodology

Research Design and Participants

In an effort to validate our previous findings, the participants in Study 3 self-selected into two competitive market positions, depending on their actual tariffs chosen. To assess the potential effects of active plan recommendations, we manipulated a between-subject factor (flat-rate bias detection: self vs. active plan recommendation). The scenario descriptions and questionnaires were pretested with a student sample (n = 20). As in Study 2, we used an online consumer panel and surveyed a sample representative of German mobile phone service customers. The experiment was conducted in 2016. In total, 308 participants completed the online experiment. They were 36 years old on average, and 48.2% of them were women.

At the beginning of the questionnaire, participants answered whether they currently used a flat-rate tariff or pay-per-use plan; participants who relied on a pay-per-use tariff structure were excluded. The remaining flat-rate tariff customers indicated the name of their service provider, their monthly billing rate, and the service provider’s perceived value proposition. We also asked them to evaluate the provider’s competitive market position (1 = definitely low cost to 7 = definitely premium). Thereafter, participants read a statement that indicated they had overspent by about 20% for their current tariff, in that their consumption pattern did not match their current allowances. Next, the participants were assigned randomly to two groups. In Group 1, participants read that they noticed their overspending by checking their monthly bill, without any communication from the service provider; in Group 2, the service provider proactively informed them about their overspending. All the other information remained constant. For the manipulation check, we used a single question about how they detected the flat-rate bias. Finally, these participants completed an online questionnaire.

Measures and Validity Evaluations

We used the same multi-item, 7-point scales from Study 2. We again assessed Cronbach’s α, and the convergent and discriminant validities of all the measures exceeded the well-established limits. Moreover, we confirmed the reliability of 2-item scales using the Spearman–Brown coefficient (Eisinga, Grotenhuis, and Pelzer 2013).

Classification of Participants According to Competitive Market Positioning

We classified participants according to information about their service provider and their own assessments of its value proposition. If the propositions included descriptions such as “high level of customer service,” they were clustered as premium providers (n = 185), whereas those described by, for example, “low-price level” indicated low-cost providers (n = 123). The competitive market position of a firm tends to be somewhat abstract for consumers who often evaluate service providers according to their value propositions and communication of those values (Bolton, Warlop, and Alba 2003). Thus, our assignment method offers the best means to capture our independent variable, namely, the firm’s strategic positioning. In line with Study 2, we coded the competitive market position as 1 for a premium provider and 0 for a low-cost provider.

To determine the adequacy of these assignments of participants to the two different competitive market positions, we evaluated consumers’ perceptions of their provider’s competitive market position (Yoo, Donthu, and Lee 2000). An analysis of variance revealed that participants’ perceptions differed significantly across the two groups. We find a significant main effect of strategic position on perceived competitive market positioning, F(1, 307) = 22.70, p < .00, M _Premium = 4.57, SD _Premium = 1.00; M _Low-cost = 2.71, SD _Low-cost = 1.19), whereas the manipulation “flat-rate bias detection” did not exert a significant main effect on perceived competitive market positioning. For greater robustness, we also evaluated consumers’ monthly billing rates in Euro using an independent samples t-test, t(301) = −9.51, p = .000; M _Premium = 33.72, SD _Premium = 17.69; M _Low-cost = 16.56, SD _Low-cost = 11.05, and cross-checked the brands against our classification.

Empirical Findings

Manipulation check

We assessed the flat-rate bias detection manipulation using a single screening question, which 12 participants failed. This failure rate is acceptable such that the flat-rate bias detection manipulation succeeded as intended.

Validation of Study 2 findings

We validate our findings from Study 2 using the same methods. The MANCOVA of internal and external attributions reveals that competitive market position has significant main effects on external attributions, F(1, 307) = 11.57, p < .00, and on internal attributions, F(1, 307) = 11.86, p < .00. Customers of premium providers with a flat-rate bias attribute their tariff choice to the provider (M = 3.59, SD = 1.65) more than customers of low-cost providers (M = 2.95, SD = 1.84). Customers of low-cost providers with a flat-rate bias instead attribute their choice to themselves (M = 5.21, SD = 1.65), compared to customers of premium providers (M = 4.64, SD = 1.45). Here again, we control for age, gender, income, and perceived competence.

To test the indirect effects, we used a multicategorical mediation analysis (Hayes 2013). The actual competitive market position of each participant’s mobile phone service provider functions as the independent variable (1 = premium provider, 0 = low-cost provider). Perceived fairness is the dependent variable, and the internal and external locus of causality are the mediating variables. The control variables remain the same. This mediation analysis again confirms the adequacy of our conceptual framework, and we validate the direct and indirect effects from Study 2 with a randomized subject allocation.

Using a 95% bootstrap CI, we determine that external attributions (effect = −.14, boot SE = .08, LLCI = −.34, ULCI = −.03) and internal attributions (effect = −.10, boot SE = .07, LLCI = −.27, ULCI = −.00) negatively mediate the relationship between the competitive market position and perceived fairness. None of the covariates significantly affects perceived fairness. Thus, Study 3 provides additional validation for our Study 2 findings; actual customers of premium service providers attribute their flat-rate bias more to the firm and display decreased fairness perceptions. The effects of a premium positioning on attributions and fairness perceptions are even stronger in this real-life setting than in the hypothetical scenario in Study 2.

Hypothesis tests

As a first step, we test the main effects of flat-rate bias detection on customer perceptions using a MANCOVA. Thereafter, we analyze the indirect effect predicted in Hypothesis 6 using PROCESS macro model 4, with customers’ flat-rate bias detection as the independent variable (1 = active plan recommendation, 0 = noticed themselves). Perceived fairness functions as the dependent variable, and the internal and external locus of causality are mediating variables. We still control for age, gender, monthly income, and perceived competence. We only expect to find an effect for premium service customers though, so we focus on them for this analysis (n = 185).

The MANCOVA of internal and external attributions reveals that flat-rate bias detection has a significant main effect on external attributions, F(1, 184) = 4.48, p < .05, but we do not find a significant main effect on internal attributions, F(1, 184) = .67, p > .10. That is, if a premium service provider fails to inform its customers about their suboptimal tariff choice, they attribute their choice to the provider (M = 3.73, SD = .14), more so than premium provider customers who were actively informed of this poor choice by the provider (M = 3.13, SD = .25).

Using the PROCESS macro with a 95% bootstrap CI, we find that external attributions (effect = −.19, boot SE = .11, LLCI = −.44, ULCI = −.03) negatively mediate the relationship between the detection of the flat-rate bias and perceived fairness in support of Hypothesis 6. If a service provider fails to inform customers about their flat-rate bias, those customers tend to attribute the bias more externally and display decreased fairness perceptions. In an additional analysis with low-cost customers, we find no direct or indirect effects of their flat-rate bias self-detection on attributions or fairness perceptions as we predicted.

Hence, given the higher value proposition in the premium segment, customers are likely to expect a higher level of consumer care in this respective market segment when compared to the low-cost segment. Thus, Study 3 highlighted that premium service providers can buffer at least some of the negative effects of customers’ external attribution by proactively issuing plan recommendations.

Overall Discussion

Theoretical Implications

Our study contributes to extant literature on flat-rate pricing in three main ways. First, our findings help explain conflicting results in choice bias literature (Ascarza and Hardie 2013; Della Vigna and Malmendier 2006; Iyengar et al. 2011; Lambrecht and Skiera 2006; Wolk and Skiera 2010; Wong 2010a, 2010b) by introducing firms’ competitive position as an important contingency factor. Two experiments show that premium providers actually suffer more from churn among flat-rate biased customers than do low-cost providers; the transactional data analysis provides initial evidence of these findings in the premium segment. The results attained with data from a premium ISP, with a relatively high price level in the market, also match existing research regarding the existence of a flat-rate bias (e.g., Della Vigna and Malmendier 2006; Nunes 2000), in that about 42% of the flat-rate customers in our sample pay too much. In this premium sample, a flat-rate bias leads to an 11.22% higher risk that customers will churn. Findings suggesting that the flat-rate bias has no negative effect on customer loyalty instead may have involved mainly low-cost contexts (Della Vigna and Malmendier 2006; Lambrecht and Skiera 2006; Wolk and Skiera 2010).

Second, this study extends flat-rate bias literature (Della Vigna and Malmendier 2006; Lambrecht and Skiera 2006; Nunes 2000; Prelec and Loewenstein 1998; Train, McFadden, and Ben-Akiva 1987) by outlining the psychological process by which a firm’s competitive position affects churn. We demonstrate that stronger external attributions of the bias reduce fairness perceptions and thus foster churn among premium provider customers. Lambrecht and Skiera (2006) stress the importance of attribution for understanding customer reactions to pricing biases, but we know of no previous literature that investigates customers’ attributions for their flat-rate biases. To fill this gap, we highlight that customers in a premium segment attribute their flat-rate bias to the provider, display decreased fairness perceptions, and churn with higher probability. Moreover, we showcase how service providers’ active tariff recommendations can reduce external attributions and thus the negative effect on perceived fairness. Customers of low-cost providers instead attribute their erroneous tariff choices internally, do not suffer from decreased fairness perceptions, and are less likely to churn.

Third, we expand and refine flat-rate bias theory (Della Vigna and Malmendier 2006; Lambrecht and Skiera 2006; Nunes 2000; Prelec and Loewenstein 1998; Train, McFadden, and Ben-Akiva 1987) by showing, with both an experiment and real-life data, that greater churn depends on the amount of monetary loss that can be attributed to the flat-rate bias. The higher price thresholds of less price-sensitive customers, together with psychological effects such as the taxi meter or insurance effect, might buffer the negative effects of the flat-rate bias on customer loyalty to some extent. Customers perceive additional value from not having to think about the cost of using a service and protecting themselves from unexpectedly high bills (Lambrecht and Skiera 2006), so they are willing to pay for those benefits. However, with each additional Euro of monetary loss, those perceived benefits diminish, which leads to customer duration decreases of almost –1%.

Practical Implications

Managers face a double bind: On the one hand, customers’ flat-rate bias is a significant profit source. According to Wong (2010b), service providers earn up to half of their revenue from rate plans that are not financially optimal for customers. On the other hand, service providers seek “zero defections” and want to enhance customer loyalty (Reichheld and Sasser 1990). Ascarza, Iyengar, and Schleicher (2016) and Min et al. (2016) note that customer retention costs service providers substantial amounts every year. To resolve this dilemma, we recommend that managers consider their own competitive position.

In the premium segment, the flat-rate bias has a negative impact on customer loyalty. Premium service providers should manage customers who exhibit high flat-rate biases proactively. For example, they could approach customers at risk and offer to switch them to a pay-per-use tariff or cheaper flat-rate offers. This procedure decreases the level of external attribution too, thereby reducing customers’ unfairness perceptions. Still, premium providers should evaluate such proactive retention campaigns carefully because they may buffer increased external attributions, but they also might decrease customer inertia and prompt customers to churn to another provider (Ascarza, Iyengar, and Schleicher 2016). Alternatively, service providers might try to increase customers’ usage levels; in the ISP context, for example, they could highlight or offer new content or complimentary video-on-demand vouchers. If customers use more data, they are less likely to experience the flat-rate bias. This option is especially pertinent for customers with a high level of flat-rate bias, who are at greater risk of churning to a competitor.

Flat-rate biased customers in the low-cost segment already are likely to stay with their provider due to their stronger internal attribution, so managers in this segment should embrace and perhaps even try to stimulate these flat-rate biases. For example, they might trigger flat-rate biases with marketing (Lambrecht and Skiera 2006) or “hedonize” their service through marketing communications and service designs that increase customers’ likelihood of selecting a flat-rate tariff (Uhrich, Schumann, and Wangenheim 2013).

Limitations and Further Research

This article offers the first investigation of the effect of the firm’s competitive market position on customer churn; it should prompt further inquiry. We included parameters directly related to the flat-rate bias in the survival analysis, so we cannot exclude the possibility of an omitted variable bias. Other factors might determine customer reactions to the flat-rate bias or the overall churn rate too, such as competitors’ advertising campaigns. Adding more parameters would be a welcome extension of our study. We also focused mainly on customer churn. Additional research might investigate consumers’ tariff switching behavior, depending on the strategic positioning and diversification of the service firm. A strong vertical product diversification might help premium providers prevent churn because customers can migrate to a cheaper alternative in the firm’s own portfolio.

In addition, in Study 2, we manipulated the level of flat-rate bias using percentage values, which lead to different monetary values depending on the service providers’ competitive market position. Yet we did not conduct an additional study, which manipulated the monetary losses in absolute values. Hence, further research could analyze whether differences in absolute amounts of monetary losses between low-cost and premium provider also lead to an increase in consumers’ churn intentions. An additional shortcoming of our article is that the transactional data study and the two experimental studies were conducted at two different points of times.

Moreover, in line with previous research, our study focuses on the flat-rate bias phenomenon, which is a frequent occurrence. Nevertheless, the so-called pay-per-use biases are evident in previous literature (Krämer and Wiewiorra 2012; Kridel, Lehman, and Weisman 1993; Lambrecht and Skiera 2006; Miravete 2002), so further research might address the impact of competitive market positions on perceptions of and behavioral reactions to the discovery of customers’ pay-per-use bias.

Finally, researchers should test different strategies that might mitigate churn rates due to customers’ flat-rate bias. In many industries, such as mobile telecommunications, companies have started to introduce pricing schemes that prevent or at least attenuate the flat-rate bias. For example, menu-based pricing models allow subscribers to customize their tariffs, according to their individual needs and consumption patterns. These tariffs also allow customers to adjust their selected options at any time, enabling them to react to the realization of their flat-rate bias immediately. Such approaches could serve as bases for scholars to derive best practices.

Supplemental Material

Supplemental Material, EXECUTIVE_SUMMARY_JSR-15-334.R3 - The Effect of a Service Provider’s Competitive Market Position on Churn Among Flat-Rate Customers

Supplemental Material, EXECUTIVE_SUMMARY_JSR-15-334.R3 for The Effect of a Service Provider’s Competitive Market Position on Churn Among Flat-Rate Customers by Sabine Moser, Jan H. Schumann, Florian von Wangenheim, Fabian Uhrich, and Felix Frank in Journal of Service Research

Footnotes

Appendix A

Table A1.

Literature Overview: Effects of Flat-Rate Bias on Customer Churn.

Author(s)	Data Set	Positive Effect of Flat-Rate Bias on Churn?
Ascarza and Hardie (2013)	Buying and renewal data of 1,173 warehouse club customers	Supported
Della Vigna and Malmendier (2006)	Gym usage data of 7,978 customers	Not supported
Iyengar, Ansari, and Gupta (2007)	Usage data of 300 mobile phone service customers	Supported
Joo, Jun, and Kim (2002)	Usage data of 6,753 telecommunication customers	Supported
Lambrecht and Skiera (2006)	Usage data of 10,882 Internet service provider customers	Not supported
Lemmens and Croux (2006)	Three data sets of mature subscribers of a major U.S. wireless telecommunications carrier	Supported
Miravete (2003)	Usage data of 1,542 telephone households	Supported
Wolk and Skiera (2010)	Usage and survey data of 941 Internet service provider customers	Not supported
Wong (2010a)	Usage data of 1,403 mobile phone service customers	Supported
Wong (2010b)	Usage data of 11,525 mobile phone service customers	Supported

Appendix B

Appendix C

Table C1.

Measurement Properties of Scales and Items (Study 2).

	Measurement Properties of Scales and Items
Scale Item	Scale Origin	α CICT	ASV FL	CR SRW	AVE IR
Locus of causality internal (1 = strongly disagree, 7 = strongly agree)	Vaidyanathan and Aggarwal (2003)	.82	.22	.78	.65
I believe that it was my own responsibility that I chose the wrong mobile phone flat-rate.		.71	.92	.92	.86
Factors external to the firm are responsible for my wrong tariff choice.		.71	.92	.91	.82
Locus of causality externally (1 = strongly disagree, 7 = strongly agree)	Vaidyanathan and Aggarwal (2003)	.92	.21	.91	.84
I think it was the service provider’s responsibility that I paid too much.		.84	.96	.89	.78
The firm is responsible for me choosing the wrong pricing plan.		.84	.96	.72	.51
Perceived fairness	Campbell (2007)	.95	.12	.88	.88
Very unfair/very fair		.92	.96	.91	.83
Wrong/right		.91	.96	.94	.88
Unreasonable/reasonable		.88	.95	.96	.92
Churn intention (1 = strongly disagree, 7 = strongly agree)	Ping (1995	.78	.11	.81	.61
I plan to churn to another mobile phone service provider.		.72	.91	.87	.76
I do not intend to stay with this provider.		.43	.91	.43	.18
In the near future, I intend to contract with another mobile phone service provider.		.72	.67	.93	.87
Perceived competence (1 = strongly disagree, 7 = strongly agree)	Fuchs, Prandelli, and Schreier (2010)	.87	.01	.87	.70
I feel competent enough to select the best mobile phone service contract.		.81	.93	.63	.43
I feel that I have the relevant knowledge and expertise to make sound evaluations.		.81	.93	.93	.86
I had difficulties evaluating the mobile phone service offers properly.		.64	.82	.92	.84

Note. α = Cronbach’s α per construct; AVE = average variance extracted per construct; CITC = corrected item-to-total correlation per item; MSV = maximum shared variance, FL = factor loading per item; CR = composite reliability; IR = indicator reliability per item; SRW = standardized regression weight per item. N = 771.

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.

Notes

References

Aggarwal

Pankaj

(2004), “The Effects of Brand Relationship Norms on Consumer Attitudes and Behavior,” Journal of Consumer Research, 31 (1), 87–101.

Andersson

Patric

Engelberg

Elisabeth

(2006), “Affective and Rational Consumer Choice Modes: The Role of Intuition, Analytical Decision-Making, and Attitudes to Money,” Working Paper , Stockholm School of Economics, Stockholm.

Ascarza

Eva

Iyengar

Raghuram

Schleicher

Martin

(2016), “The Perils of Proactive Churn Prevention Using Plan Recommendations: Evidence from a Field Experiment,” Journal of Marketing Research, 53 (1), 46–60.

Ascarza

Eva

Hardie

Bruce G. S.

(2013), “A Joint Model of Usage and Churn in Contractual Settings,” Marketing Science, 32 (4), 570–590.

Ascarza

Eva

Lambrecht

Anja

Vilcassim

Naufel

(2012), “When Talk Is “Free”: The Effect of Tariff Structure on Usage Under Two- and Three-Part Tariffs,” Journal of Marketing Research, 49 (6), 882–899.

Bitner

Mary J.

(1990), “Evaluating Service Encounters: The Effects of Physical Surroundings and Employee Responses,” Journal of Marketing, 54 (2), 69–82.

Bolton

Lisa E.

Warlop

Luk

Alba

Joseph W.

(2003), “Consumer Perceptions of Price (Un)Fairness,” Journal of Consumer Research, 29 (4), 474–491.

Boshoff

Christo

Leong

Jason

(1998), “Empowerment, Attribution and Apologising as Dimensions of Service Recovery,” International Journal of Service Industry Management, 9 (1), 24–47.

Brislin

Richard W.

(1970), “Back-Translation for Cross-Cultural Research,” Journal of Cross-Cultural Psychology, 1 (3), 185–216.

10.

Campbell

Margaret C.

(2007), ““Says Who?!” How the Source of Price Information and Affect Influence Perceived Price (Un)Fairness,” Journal of Marketing Research, 44 (2), 261–271.

11.

Choi

Seon-Kyou

Lee

Myeong-Ho

Chung

Gyu-Hwa

(2001), “Competition in Korean Mobile Telecommunications Market: Business Strategy and Regulatory Environment,” Telecommunications Policy, 25 (1/2), 125–138.

12.

Dabholkar

Pratibha A.

Shepherd

C. D.

Thorpe

Dayle I.

(2000), “A Comprehensive Framework for Service Quality: An Investigation of Critical Conceptual and Measurement Issues Through a Longitudinal Study,” Journal of Retailing, 76 (2), 139–173.

13.

Day

George S.

Nedungadi

Prakash

(1994), “Managerial Representations of Competitive Advantage,” Journal of Marketing, 58 (2), 31–44.

14.

Della Vigna

Stefano

Malmendier

Ulrike

(2006), “Paying Not to Go to the Gym,” The American Economic Review, 96 (3), 694–719.

15.

Dess

Gregory G.

Davis

Peter S.

(1984), “Porter’s (1980) Generic Strategies as Determinants of Strategic Group Membership and Organizational Performance,” Academy of Management Journal, 27 (3), 467–488.

16.

Eisinga

Rob

Grotenhuis

Manfred t.

Pelzer

Ben

(2013), “The Reliability of a Two-item Scale: Pearson, Cronbach, or Spearman-Brown?,” International Journal of Public Health, 58 (4), 1–6.

17.

Folkes

Valerie S.

(1988), “Recent Attribution Research in Consumer Behavior: A Review and New Directions,” Journal of Consumer Research, 14 (4), 548–565.

18.

Folkes

Valerie S.

Koletsky

Susan

Graham

John L.

(1987), “A Field Study of Causal Inferences and Consumer Reaction: The View from the Airport,” Journal of Consumer Research, 13 (4), 534–539.

19.

Folkes

Valerie S.

(1984), “Consumer Reactions to Product Failure: An Attributional Approach,” Journal of Consumer Research, 10 (4), 398–409.

20.

Fornell

Claes

Larcker

David F.

(1981), “Evaluating Structural Equation Models with Unobservable Variables and Measurement Error,” Journal of Marketing Research, 18 (1), 39–50.

21.

Fuchs

Christoph

Prandelli

Emanuela

Schreier

Martin

(2010), “The Psychological Effects of Empowerment Strategies on Consumers’ Product Demand,” Journal of Marketing, 74 (1), 65–79.

22.

Grønnesby

Jon K.

Borgan

Ørnulf

(1996), “A Method for Checking Regression Models in Survival Analysis Based on the Risk Score,” Lifetime Data Analysis, 2 (4), 315–328.

23.

Han

Sangman

Gupta

Sunil

Lehmann

Donald R.

(2001), “Consumer Price Sensitivity and Price Thresholds,” Journal of Retailing, 77 (4), 435–456.

24.

Haws

Kelly L.

Bearden

William O.

(2006), “Dynamic Pricing and Consumer Fairness Perceptions,” Journal of Consumer Research, 33 (3), 304–311.

25.

Hayes

Andrew F.

(2013), Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. New York, NY: Guilford Press.

26.

Heidenreich

Sven

Handrich

Matthias

(2010), “The Fascination of Limitless Consumption – An Empirical Study of Existence and Causes of Flat-Rate Biases in the Cellular Mobile Industry,” Global Marketing Conference, Tokyo, Japan.

27.

Heider

Fritz

(1958), The Psychology of Interpersonal Relations. New York: Wiley.

28.

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.

29.

Hess

Ronald L.

Ganesan

Shankar

Klein

Noreen M.

(2007), “Interactional Service Failures in a Pseudorelationship: The Role of Organizational Attributions,” Journal of Retailing, 83 (1), 79–95.

30.

Hill

Charles W. L.

(1988), “Differentiation Versus Low Cost or Differentiation and Low Cost: A Contingency Framework,” The Academy of Management Review, 13 (3), 401–412.

31.

Hobson

Michael

Spady

Richard H.

(1988), “The Demand for Local Telephone Service Under Optional Local Measured Service,” Bellcore Economics Discussion Paper No. 50, Morristown, NJ.

32.

Homburg

Christian

Workman

John P.

Krohmer

Harley

(1999), “Marketing’s Influence Within the Firm,” Journal of Marketing, 63 (2), 1–17.

33.

Hosmer

David W.

Lemeshow

Stanley

May

Susanne

(2008), Applied Survival Analysis: Regression Modeling of Time-to-Event Data, 2nd ed. Hoboken, NJ: Wiley-Interscience.

34.

Iyengar

Raghuram

Jedidi

Kamel

Essegaier

Skander

Danaher

Peter J.

(2011), “The Impact of Tariff Structure on Customer Retention, Usage, and Profitability of Access Services,” Marketing Science, 30 (5), 820–836.

35.

Iyengar

Raghuram

Ansari

Asim

Gupta

Sunil

(2007), “A Model of Consumer Learning for Service Quality and Usage,” Journal of Marketing Research, 44 (4), 529–544.

36.

Jamal

Zainab

Bucklin

Randolph E.

(2006), “Improving the Diagnosis and Prediction of Customer Churn: A Heterogeneous Hazard Modeling Approach,” Journal of Interactive Marketing, 20 (3-4), 16–29.

37.

Joo

Young-Hyuck

Jun

Jong-Kun

Kim

Byung-Do

(2002), “Encouraging Customers to Pay Less for Mobile Telecommunication Services,” Journal of Database Marketing, 9 (4), 350–359.

38.

Kalafatis

Stavros P.

Tsogas

Markos H.

Blankson

Charles

(2000), “Positioning Strategies in Business Markets,” Journal of Business & Industrial Marketing, 15 (6), 416–437.

39.

Kalra

Ajay

Goodstein

Ronald C.

(1998), “The Impact of Advertising Positioning Strategies on Consumer Price Sensitivity,” Journal of Marketing Research, 35 (2), 210–224.

40.

Kelley

Harold H.

(1967), “Attribution Theory in Social Psychology,” Nebraska Symposium on Motivation, 15, 192–238.

41.

Krämer

Jan

Wiewiorra

Lukas

(2012), “Beyond the Flat Rate Bias: The Flexibility Effect in Tariff Choice,” Telecommunications Policy, 36 (1), 29–39.

42.

Kridel

Donald J.

Lehman

Dale E.

Weisman

Dennis L.

(1993), “Option Value, Telecommunications Demand, and Policy,” Information Economics and Policy, 5 (2), 125–144.

43.

Lambrecht

Anja

Skiera

Bernd

(2006), “Paying Too Much and Being Happy About It: Existence, Causes, and Consequences of Tariff-Choice Biases,” Journal of Marketing Research, 43 (2), 212–223.

44.

Lemmens

Aurélie

Croux

Christophe

(2006), “Bagging and Boosting Classification Trees to Predict Churn,” Journal of Marketing Research, 43 (2), 276–286.

45.

Shaomin

(1995), “Survival Analysis: Just as in Life, Customers Are ‘Born’ and They ‘Die’. Here’s How to Prevent Their Premature Demise,” Marketing Research, 7 (4), 16–23.

46.

May

Susanne

Hosmer

David W.

(1998), “A Simplified Method of Calculating an Overall Goodness-of-Fit Test for the Cox Proportional Hazards Model,” Lifetime Data Analysis, 4 (2), 109–120.

47.

McAlister

Leigh

Srinivasan

Raji

Jindal

Niket

Cannella

Albert A.

(2016), “Advertising Effectiveness: The Moderating Effect of Firm Strategy,” Journal of Marketing Research, 53 (2), 207–224.

48.

McKee

Daryl O.

Varadarajan

P. R.

Pride

William M.

(1989), “Strategic Adaptability and Firm Performance: A Market-Contingent Perspective,” Journal of Marketing, 53 (3), 21–35.

49.

Mehta

Nitin

Rajiv

Surendra

Srinivasan

Kannan

(2003), “Price Uncertainty and Consumer Search: A Structural Model of Consideration Set Formation,” Marketing Science, 22 (1), 58–84.

50.

Min

Sungwook

Zhang

Xubing

Kim

Namwoon

Srivastava

Rajendra K.

(2016), “Customer Acquisition and Retention Spending: An Analytical Model and Empirical Investigation in Wireless Telecommunications Markets,” Journal of Marketing Research, 53 (5), 728–744.

51.

Miravete

Eugenio J.

(2003), “Choosing the Wrong Calling Plan? Ignorance and Learning,” American Economic Review, 93 (1), 297–310.

52.

Miravete

Eugenio J.

(2002), “Estimating Demand for Local Telephone Service With Asymmetric Information and Optional Calling Plans,” Review of Economic Studies, 69 (4), 943–971.

53.

Mitchell

Bridger M.

Vogelsang

Ingo

(1991), Telecommunications Pricing: Theory and Practice. Cambridge, MA: Cambridge University Press.

54.

Monroe

Kent B.

(1973), “Buyers’ Subjective Perceptions of Price,” Journal of Marketing Research, 10 (1), 70–80.

55.

Murray

Alan I.

(1988), “A Contingency View of Porter’s “Generic Strategies”,” The Academy of Management Review, 13 (3), 390–400.

56.

Nunes

Joseph C.

(2000), “A Cognitive Model of People’s Usage Estimations,” Journal of Marketing Research, 37 (4), 397–409.

57.

Oliver

Richard L.

(2010), Satisfaction: A Behavioral Perspective on the Consumer. New York: ME Sharpe.

58.

Peterson

Christopher

Semmel

Amy

Baeyer

Carl von

Abramson

Lyn Y.

Metalsky

Gerald I.

Seligman

Martin E. P.

(1982), “The Attributional Style Questionnaire,” Cognitive Therapy and Research, 6 (3), 287–299.

59.

Pick

Doreén

Thomas

Jacquelyn S.

Tillmanns

Sebastian

Krafft

Manfred

(2016), “Customer Win-Back: The Role of Attributions and Perceptions in Customers’ Willingness to Return,” Journal of the Academy of Marketing Science, 44 (2), 218–240.

60.

Ping

Robert A.

(1995), “Some Uninvestigated Antecedents of Retailer Exit Intention,” Journal of Business Research, 34 (3), 171–180.

61.

Popkowski Leszczyc

Peter T. L.

Rao

Ram C.

(1990), “An Empirical Analysis of National and Local Advertising Effect on Price Elasticity,” Marketing Letters, 1 (2), 149–160.

62.

Porter

Michael E.

(1996), “What Is Strategy?,” Harvard Business Review, 74 (6), 61–78.

63.

Porter

Michael E.

(1980), Competitive Strategy: Techniques for Analyzing Industries and Competitors, 52. printing. New York: Free Press.

64.

Preacher

Kristopher J.

Hayes

Andrew F.

(2008), “Asymptotic and Resampling Strategies for Assessing and Comparing Indirect Effects in Multiple Mediator Models,” Behavior Research Methods, 40 (3), 879–891.

65.

Prelec

Drazen

Loewenstein

George

(1998), “The Red and the Black: Mental Accounting of Savings and Debt,” Marketing Science, 17 (1), 4–28.

66.

Reichheld

Frederick F.

Sasser

W. E.

(1990), “Zero Defections: Quality Comes to Services,” Harvard Business Review, 68 (5), 105–111.

67.

Russell

Dan

(1982), “The Causal Dimension Scale: A Measure of How Individuals Perceive Causes,” Journal of Personality and Social Psychology, 42 (6), 1137–1145.

68.

Santonen

Teemu

(2007), “Price Sensitivity as an Indicator of Customer Defection in Retail Banking,” International Journal of Bank Marketing, 25 (1), 39–55.

69.

Scott

Maura L.

Mende

Martin

Bolton

Lisa E.

(2013), “Judging the Book by Its Cover? How Consumers Decode Conspicuous Consumption Cues in Buyer-Seller Relationships,” Journal of Marketing Research, 50 (3), 334–347.

70.

Train

Kenneth E.

(1991), Optimal Regulation. The Economic Theory of Natural Monopoly. Cambridge: MIT Press.

71.

Train

Kenneth E.

McFadden

Daniel L.

Ben-Akiva

Moshe

(1987), “The Demand for Local Telephone Service: A Fully Discrete Model of Residential Calling Patterns and Service Choices,” The RAND Journal of Economics, 18 (1), 109–123.

72.

Tsiros

Michael

Mittal

Vikas

Ross

William T.

(2004), “The Role of Attributions in Customer Satisfaction: A Reexamination,” Journal of Consumer Research, 31 (2), 476–483.

73.

Uhrich

Fabian

Schumann

Jan H.

Wangenheim

Florian von

(2013), “The Impact of Consumption Goals on Flat-Rate Choice: Can “Hedonizing” a Service Increase Customers’ Propensity to Choose a Flat Rate?,” Journal of Service Research, 16 (2), 216–230.

74.

Vaidyanathan

Rajiv

Aggarwal

Praveen

(2003), “Who Is the Fairest of Them All? An Attributional Approach to Price Fairness Perceptions,” Journal of Business Research, 56 (6), 453–463.

75.

Van Vaerenbergh

Yves

Orsingher

Chiara

Vermeir

Iris

Lariviere

Bart

(2014), “A Meta-Analysis of Relationships Linking Service Failure Attributions to Customer Outcomes” Journal of Service Research, 17 (4), 381–398.

76.

Weiner

Bernard

(1985), “An Attributional Theory of Achievement Motivation and Emotion,” Psychological Review, 92 (4), 548–573.

77.

Weiner

Bernard

(1986), “An Attributional Theory of Achievement Motivation and Emotion,” in An Attributional Theory of Motivation and Emotion, Weiner

, ed. New York: Springer, 159–190.

78.

Wolk

Agnieszka

Skiera

Bernd

(2010), “Tariff-Specific Preferences and Their Influence on Price Sensitivity,” Business Research, 3 (1), 70–80.

79.

Wong

Ken K.-K.

(2010a), “Fighting Churn with Rate Plan Right-Sizing: A Customer Retention Strategy for the Wireless Telecommunications Industry,” The Service Industries Journal, 30 (13), 2261–2271.

80.

Wong

Ken K.-K.

(2010b), “Should Wireless Carriers Protect Residential Voice Subscribers from High Overage and Underage Charges? Insights from the Canadian Telecommunications Market,” Telecommunications Policy, 34 (8), 461–481.

81.

Xia

Lan

Monroe

Kent B.

Cox

Jennifer L.

(2004), “The Price Is Unfair! A Conceptual Framework of Price Fairness Perceptions,” Journal of Marketing, 68 (4), 1–15.

82.

Yoo

Boonghee

Donthu

Naveen

Lee

Sungho

(2000), “An Examination of Selected Marketing Mix Elements and Brand Equity,” Journal of the Academy of Marketing Science, 28 (2), 195–211.

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

0.01 MB