Rethinking Picky Shoppers and Store Reputation: Effective Online Service Recovery Strategies for Products with Minor Defects

Abstract

This study examines how the initiation of monetary compensation and the sincerity of apologies influence customers’ repurchase intentions after receiving products with minor defects, and how these effects are moderated by consumer pickiness and the e-store’s reputation. Two scenario-based, between-subjects online experiments were conducted in China via the Wenjuanxing online survey platform, with participants voluntarily taking part in exchange for monetary compensation. We propose that consumers have a higher repurchase intention when online customer service initiatively offers monetary compensation (vs. passive vs. no compensation) and that the effect of monetary compensation modes can be enhanced by a serious apology from customer service. Building on these findings, Study 1 (n = 330) reveals that picky consumers are more sensitive to the enhanced effect of initiative monetary compensation on repurchase intentions under serious apologies. Moreover, Study 2 (n = 537) confirms that consumers are more stringent in their expectations of high-reputation stores, whereas low-reputation stores can achieve similar positive effects with casual apologies that high-reputation stores only obtain through serious apologies. This study provides theoretical support for the importance of the initiation of actions and sincerity in service recovery and offers guidance for businesses dealing with picky consumers.

1. Introduction

The virtual nature of online shopping and the complexities of the delivery process contribute to the inevitability of online service failures [1]. Service failures and service recovery have long been regarded as important topics in both academic research and managerial practice within the global marketplace. In the U.S. context, Jung and Seock [2] examined how apology, compensation, and their combination in online shopping influence customer behavior. In Europe, Abeler et al. [3] ran a field experiment on German eBay, comparing apologies with cash compensation of varying amounts for inducing withdrawal of negative feedback. This issue is likewise salient in Asian markets. Notably, compared with consumers in Europe and North America, Asian consumers place greater emphasis on relational harmony [4], making research on service recovery in Asia both timely and crucial.

Research into the classifications of e-commerce service failures reveals that product defects constitute a significant portion, exceeding 20% of the reported issues [5]. Upon the delivery of a product to a customer, diverse defects may manifest under unforeseen circumstances that are not expected by the seller [6]. Severe product defects have the potential to result in personal injury, while minor defects tend to pertain primarily to cosmetic concerns. The presence of severe product defects can evoke more negative emotional responses [7]. If the initial problem is serious, even after receiving multiple apologies and compensation from the company, the customers may still feel uneasy, generating negative word of mouth and making it less likely for the seller to build trust and commitment [8].

Thus, service recovery is more effective if the product defects experienced are not serious and are within tolerance limits [9]. Most studies related to online service recovery have concentrated on scenarios involving delivery, logistics, packaging, or product quality issues [2,10,11,12], which are the primary concerns for online consumers [13]. These defects, often referred to as “minor defects,” are perceived in the product and are generally located in non-central areas, primarily affecting aesthetics and appearance without significantly impacting product usability or functionality [14]. Considering the temporal cost for the return and exchange processes, when consumers receive such products with minor defects, they prefer seeking quicker and more effortless solutions by consulting customer service, as the perceived defects are outweighed by other expected attributes of the product [15]. Ignoring or misinterpreting customer needs at this point could easily increase consumer regret, provoke anger, and escalate minor issues [16,17].

Economic (e.g., monetary compensation) and psychological (e.g., apology) remedies are the two broad categories of service recovery strategies [18]. However, the realm of online shopping service failures and subsequent recovery has remained relatively unexplored, yielding only a sparse and contradictory array of evidence regarding the efficacy of these dual remedial factors [2,10,11,12,19,20,21,22,23] (see

Table 1. Review of the literature on online service failure, apology, and compensation.

Study	Online Service Failure Type	More Effective Strategy	Key Findings
Fan and Zheng [20]	Slow delivery	Compensation	Compensation boosts repurchase after service failures. Apologies alone do not suffice for satisfaction.
Jung and Seock [2]	Slow delivery and poor communication	Apology	Apologies enhance perceived fairness and interactional justice, with or without compensation.
Jafarzadeh et al. [21]	Packaging errors	Compensation	Economic compensation is most effective, signaling a high recovery effort.
Ma and Zhong [10]	Inferior product	Apology	Unethical service failures prompt a preference for apologies; ethical services benefit from both.
Wei and Lin [12]	Obvious product quality	Apology	Apologies outperform compensation in fostering positive emotions and diminishing negative ones.
Wei et al. [23]	Wrong and used product	Apology	Both compensation and apologies promote forgiveness; apologies do so more effectively.
Noor et al. [22]	Product falls below expectation	Compensation	The impact of compensation on the willingness to purchase is significant, whereas the impact of apologies is not.
Roy et al. [11]	Logistical issues	Compensation	Apologies alone are not effective in establishing justice to achieve service recovery.
Azizi et al. [19]	Unknown extra fees	Apology and compensation	Combining apology and compensation has the greatest impact on repatronage intentions.

). Consequently, these findings lack conclusiveness when it comes to prognosticating consumers’ reactions to service recovery for minor imperfections. Furthermore, prior investigations have primarily appraised the mere presence or absence of compensation and apology, overlooking the magnitudes of them. This has led numerous enterprises to rigidly apply uniform recovery approaches, when in fact the deployment of tailored and effective service strategies is pivotal for genuinely reinstating customer prestige and repurchase intentions [24,25].

To devise customized service recovery strategies for minor-defect products, understanding consumer characteristics is crucial. Picky consumers have always been both a nightmare and a source of annoyance for merchants due to their narrow and unpredictable preferences [26]. Furthermore, picky consumers are prone to detecting minor flaws in products, leading to a higher intention for returns [27]. However, this conclusion was reached in the context of no service recovery being provided. Current research has not yet provided an answer to the question of how to win back picky customers, and we attempt to fill this research gap.

Store reputation is also one of the important factors affecting the effectiveness of service remediation. Online consumers often face challenges in accurately assessing product quality based on intrinsic characteristics, leading them to use seller reputation as a crucial cue [28]. Su and Tippins [29] indicated that a high reputation can be a less plausible reason for product failure when compared to lower-reputation stimuli, which tend to set lower expectations. Conversely, when high-reputation sellers provide service recovery, consumer forgiveness has been proven to exert a more pronounced positive influence on repurchase intentions compared to low-reputation sellers [23]. Existing research on reputation and service recovery has yielded mixed results, and limited attention has been paid to the role of reputation in the context of service recovery for products with minor defects. This study also aims to fill this gap in the literature.

Therefore, the primary research focus of this study is twofold: RQ1. How do service recovery strategies for products with minor flaws impact customer repurchase intentions? RQ2. What are the roles of consumers’ pickiness and sellers’ reputations as moderating factors in this relationship? Additionally, the concepts of compensation and apology are dissected according to varying degrees of initiative and sincerity. This approach is used to develop the most detailed and cost-effective service recovery templates for minor defects, enriching the application scenarios of online service recovery.

2. Theoretical Background and Hypotheses Development

2.1. The Initiation of Monetary Compensation in Service Recovery

Service recovery refers to the effort of addressing issues encountered in service delivery before customers lodge complaints or before they depart dissatisfied with their service experience [30]. Prior research has substantiated a direct and positive correlation between effective service recovery and customers’ willingness to repurchase [31].

Monetary compensation is a powerful service remediation measure, possessing characteristics of certainty, tangibility, and observability [32]. Acquiring financial compensation can effectively counterbalance losses suffered by consumers owing to service failure, fostering the perception that the company is compassionate in addressing concerns [33]. A substantial portion of earlier research has centered around the size of compensation [34] or the availability of monetary restitution [20].

However, the way a company initiates monetary compensation also has an impact on the effectiveness of service recovery. Proactive initiation refers to the company initiating contact with the customer on its own, whereas passive initiation occurs only after the customer seeks assistance. Research indicates that having customers take on the role of recovery initiators results in lower average recovery ratings and reduced customer retention rates [35]. Organization-initiated recoveries allow customers to perceive a higher level of interactional justice compared to customer-initiated recoveries [36]. However, despite evidence indicating that customers prefer service providers offering proactive interactions [37], the focus of service marketing research remains predominantly on passive service recovery concerns rather than delving deeper into an understanding of proactive service failure prevention strategies [38]. Consequently, in the realm of service recovery, we assert that seller-initiated financial compensation is more likely to enhance customer repurchase intention compared to passive recovery or no compensation.

H1:

When online store customer service takes the initiative to provide monetary compensation (vs. passive vs. no) to customers, the customers’ intention to repurchase will be higher.

2.2. Apology Sincerity in Service Recovery

Corporate apologies can convey courtesy, concern, and effort to consumers following service failures, thus elevating their assessment of the company [39]. According to the social exchange theory, apologies reallocate respect as a critical social resource within the exchange relationship [40]. By offering emotional recovery services, businesses signal to consumers their willingness to take responsibility for failures and express concern for their feelings [33,41].

Simply providing financial compensation might not be adequate to shift consumer reactions from negative to positive. Bitner et al. [42] caution that offering compensation without proper explanation could lead consumers to perceive the company as guilty, ultimately resulting in negative evaluations. This is because apologies signify respect for consumers, countering the sense of injustice caused by service failures and showcasing the company’s sense of responsibility. Conversely, monetary compensation alone could heighten consumer suspicion regarding the service provider’s deliberate service failure [10]. Moreover, engaging in complaining behavior often entails psychological costs; customers can feel embarrassed and personally distressed when anticipating unwelcome or unpleasant confrontations with employees [43]. Consequently, the presence of an apology can mitigate the psychological costs borne by customers and signify the effect of monetary compensation. Similarly, apologies entailing economic costs are more adept at facilitating forgiveness than cost-free apologies are [44].

However, since the effectiveness of an apology comes from the company respecting its customers and accepting responsibility for service failures [27,35], the moderating effect of an apology depends on its sincerity. A sincere apology holds greater potential to elicit forgiveness from individuals than an insincere one [45]. A sincere apology implies that the apologizing organization authentically values people’s emotions and acknowledges the hardships faced by victims [46]. A deeply sincere apology proves more effective in bolstering consumers’ future purchase intentions compared to a superficial apology lacking sincerity [47].

Inauthentic apologies can backfire, potentially casting the wrongdoer as untrustworthy, self-serving, and harboring ulterior motives [48]. When addressing customer dissatisfaction, insincere apologies prove both ineffectual and vacuous, as they might be perceived as manipulative endeavors to pacify the victim’s adverse emotions [49]. In other words, monetary compensation with insincere apologies would fail in earning back customers’ forgiveness and trust. From this, we deduce that a sincere apology can enhance the positive effect of compensation on customers’ repurchase intentions, whereas casual apologies or the absence of an apology fail to yield such outcomes. Therefore, we can hypothesize the following:

H2:

The effect of monetary compensation initiation on repurchase intention is moderated by apology sincerity.

H2-1:

When a customer receives a serious apology, monetary compensation initiation has a significant effect on repurchase intention.

H2-2:

When a customer receives a casual or no apology, monetary compensation initiation does not have a significant effect on repurchase intention.

2.3. The Constructive Side of Consumer Pickiness

Picky consumers are defined as those who tend to set clear, specific, and strict purchasing standards during the shopping process and only accept products that closely match their personal preferences [27]. For consumers who are picky about shopping, their acceptance range for products and stores is very narrow, making it hard to make purchasing decisions [50]. Furthermore, picky consumers have been demonstrated to correlate with high self-confidence [27]. They rely more heavily on their own personal experiences and independent sources of information during the decision-making process [51], and they demonstrate an “excessive” tendency in their selection and decision-making behaviors, often characterized by overly meticulous and idiosyncratic criteria [26,27]. This often leads them to invest significant time and effort in selecting stores [51]. Such traits make them highly sensitive to service failures. Studies show that people often reject feedback that conflicts with their self-perception to maintain their confidence [52].

When service failures occur, we speculate that minor product defects may undermine picky consumers’ confidence in their purchase decisions. Given their heightened attention to detail [27], we assume that only sincere and serious apologies from sellers are likely to make these consumers feel respected, helping them attribute the failure to the company rather than themselves, thereby alleviating their negative emotions. Accordingly, we hypothesize that only serious apologies can enhance the effectiveness of monetary compensation in service recovery for picky consumers. In contrast, less picky consumers may approach shopping with greater openness and flexibility. When they are dissatisfied with a purchase, they are more likely to switch to another store. Because these consumers are less concerned with fine details, even a casual apology from the seller may have a positive effect.

Previous research has predominantly focused on the negative aspects of picky consumers [27,50]. However, the findings of this study suggest that if you can capture the hearts of picky consumers, they have the potential to become the most loyal customers. Thus, we can deduce that in scenarios where service recoveries are perfect, low-picky consumers may exhibit a paradoxically lower propensity to repurchase compared to their pickier counterparts.

H3:

The moderating effect of apology sincerity (H2) is stronger among high-pickiness customers than among low-pickiness customers.

H3-1:

For high-pickiness customers, apology sincerity strengthens the effect of the monetary compensation mode on the customers’ repurchase intention.

H3-2:

For low-pickiness customers, apology sincerity does not significantly strengthen the effect of the monetary compensation mode on the customers’ repurchase intention.

2.4. E-Store Reputation as Expectation and Pressure

Unlike traditional offline transactions, in the process of online shopping, consumers cannot directly touch or try products to assess their quality, and it is challenging to ensure the reliability of sellers or products [53]. Consequently, consumers face more pronounced transaction risks stemming from information asymmetry compared to offline transactions [54].

Seller reputation has been unequivocally shown to wield a pivotal role in the realm of online shopping. Consumers modify their bidding behavior to accommodate surcharges when purchasing from sellers with a low reputation, whereas they do not make such adjustments when dealing with high-reputation sellers. Also, when consumers opt to procure products from sellers with lower reputations, they expend a more protracted duration in making their purchase decisions [55]. Xu et al. [56] discovered that consumers are even willing to pay a premium for a seller’s high reputation.

A good reputation is a double-edged sword. The expectancy disconfirmation theory suggests that customers form expectations about service performance before they make a purchase [57], and highly reputable sellers often benefit from an initial positive impression. Consequently, when encountering favorable product information associated with these sellers, consumers are more inclined to reinforce their existing trust and identification. Consumers inherently hold elevated expectations for highly reputable sellers, presuming that their product quality surpasses that of sellers with lower reputations [58]. If products from reputable sellers happen to be flawed, these expectations are not met, disappointment follows [59], and the negative anticipation may send strong signals indicating a disparity between the actual product and the seller’s ratings, implying subpar product quality, leading to a sense of shock and betrayal [60]. In contrast, when dealing with low-reputation sellers, product defects might not trigger the same level of anger; given that consumers have already adjusted their expectations downward for these sellers, product issues merely validate their already modest expectations for sellers with low reputations [28].

Likewise, within the context of service failures, customers expect highly reputable sellers to offer high-quality service recovery measures that demonstrate their competence. This involves not only proactive and meaningful financial compensation but also apologies. Merely offering an apology does not demonstrate that the employees or companies have genuinely put in effort [61]; they must also apologize sincerely. Conversely, for low-reputation sellers, customers perceive their products as already deserving low ratings. They do not anticipate exceptional service and are content if the seller can economically address the product’s shortcomings. Thus, if an apology is given, regardless of its sincerity, it can satisfy them. Accordingly, we propose the following hypothesis:

H4:

An interaction exists among monetary compensation initiation, apology sincerity, and reputation on repurchase intention.

H4-1:

For a high-reputation online store, a customer’s repurchase intention is higher when they receive initiative (vs. passive vs. no) monetary compensation and a serious apology, whereas such a difference in repurchase intention is not observed with a casual or no apology.

H4-2:

For a low-reputation online store, a customer’s repurchase intention is higher when they receive initiative (vs. passive vs. no) monetary compensation and a serious or casual apology, whereas such a difference in repurchase intention is not observed with no apology.

The relations among constructs in the above hypotheses are summarized in Figure 1.

3. Methods

We tested our hypotheses through two scenario-based online experiments. Data were collected via Wenjuanxing, one of the most widely used online survey platforms in China, functionally comparable to MTurk or Qualtrics. All participants were clearly informed of the academic purpose and background of this study prior to participation and voluntarily consented to take part in the survey. The questionnaire remained open on the platform for two weeks, and participants who completed it received a monetary reward. The final sample size meets the commonly recommended thresholds in prior research for experimental designs [62], ensuring sufficient statistical power and robust analysis. As shown in

Table 2. Demographic characteristics.

Items	Value	Percentage (%)
		Study 1 (n = 330)	Study 2 (n = 537)
Gender	Male	48.5	47.1
	Female	51.5	52.9
Age	18–20	9.7	4.7
	20 s	20.0	28.5
	30 s	43.6	43.4
	40 s	22.7	20.5
	>50	3.9	3.0
Education	Below high school	1.5	3.2
	High school degree	26.1	9.7
	Bachelor’s degree	60.3	61.3
	Master’s degree and above	12.1	25.9
Monthly income (RMB)	<1000	9.7	0.9
	1001–3000	1.8	3.7
	3001–5000	32.1	20.3
	5001–10,000	42.7	48.0
	>10,000	13.6	27.0

, participants in both experiments were predominantly in their 30 s, held a bachelor’s degree, and had a monthly income ranging from RMB 5001 to 10,000, which reflects the characteristics of the most active consumer group in the e-commerce sector.

3.1. Study 1

Study 1 served three objectives. First, it aimed to validate H1, which states customers are more likely to repurchase from an online store when the seller initiatively offers compensation compared to when compensation is passive or not provided. Second, Study 1 aimed to test the moderating effect of apology sincerity on the relationship between the initiative of monetary compensation and customers’ repurchase intentions (H2). Finally, Study 1 also provided evidence for the moderated moderating effect of consumers’ pickiness level in everyday shopping (H3).

3.1.1. Design and Participants

Study 1 used a 3 (monetary compensation: initiative vs. passive vs. no) × 3 (apology: serious vs. casual vs. no) between-subjects design. We recruited 330 volunteers (51.52% female) in China using the Wenjuanxing platform in exchange for monetary compensation. The demographic characteristics of the sample are summarized in Table 2.

The stimulus materials for the service failure and service recovery types were composed of two parts, including written materials and screenshots that provided a conversation between the consumer and the e-commerce platform’s online customer service. The design of our minor defective product service failure followed a design used in a prior study [63]. All participants read a similar introduction to the service failure scenario: “Please imagine that you purchased a badminton racket online. A few days later the product arrived. Unfortunately, you found that it was a defective product—there was a scratch on the product, although this defect did not affect product performance. You contacted the e-store’s online customer service and reported the flaw in this product”. Then the subjects viewed one of nine randomly assigned chatting screenshots (see Appendix A) between a customer and online customer service. The monetary compensation given by the customer service was either in the initiative-, passive-, or no-compensation mode. The apology conditions were either in the serious-, casual-, or no-apology mode, with the apology content based on previous research [64]. Participants in the sincere-apology group saw the apology “I am sorry for the inconvenience,” while the casual-apology group received “Sorry for the mishap.”

Next, participants rated repurchase intentions with two items [10] (Cronbach’s α = 0.95; see Appendix C). Then, we measured participants’ levels of pickiness with eight items [27] (Cronbach’s α = 0.88). Additionally, as a manipulation check, participants rated their agreement on the type of initiation of monetary compensation (initiative, passive, no) on a 7-point Likert scale [36]. Two 7-point Likert-type items measured perceived apology sincerity [65] (Cronbach’s α = 0.96). Customers’ perceived severity of a product defect was measured on a 5-point scale [9]. Ratings between 1 and 3 on the severity scale were perceived as low-severity product defects, whereas scores 4 and 5 were perceived as high-severity [9]. Finally, participants indicated demographics before ending the procedure (see Table 2).

The severity of service failure due to the product defect of a badminton racket having scratches was analyzed with a one-sample t-test, revealing a mean severity rating of 2.28, which significantly differs from the center point (3) of the scale (t(329) = −12.87, p < 0.001), classifying it within the low-severity range [9]. One-way ANOVA and post hoc analysis results showed significant differences in the level of initiative among the groups with no, passive, and initiative monetary compensation as Mnmc (no compensation) = 1.50, SD = 0.88 vs. Mpmc (passive compensation) = 2.30, SD = 1.40 vs. Mimc (initiative compensation) = 5.62, SD = 1.59, F(2, 327) = 299.04, all ps < 0.001. Differences in the sincerity of the apology were also significant among no-apology, casual-apology, and serious-apology groups as Mnap (no apology) = 1.77, SD = 0.81 vs. Mcap (casual apology)= 2.29, SD = 1.17 vs. Msap (serious apology) = 5.46, SD = 1.43, F(2, 327) = 327.50, all ps < 0.001. These results suggest successful manipulation of product defect severity, monetary compensation, and apology sincerity.

3.1.2. Results

We employed a two-way ANOVA to test the interaction effect of monetary compensation and apology sincerity on repurchase intention. The results revealed a main effect of monetary compensation on customer repurchase intention (F(2, 321) = 52.71, p < 0.001, η_p² = 0.25). The intention to repurchase was higher in consumers when online customer service initiatively offered monetary compensation (Mimc = 3.49, SD = 0.09) compared to the passive group (Mpmc = 2.86, SD = 0.09, p < 0.001) and the no-monetary-compensation group (Mnmc = 2.23, SD = 0.09, p < 0.001), supporting H1. The main effect of apology sincerity on repurchase intention was also significant (F(2, 321) = 343.99, p < 0.001, η_p² = 0.68). Consistent with expectations, there was a significant interaction between monetary compensation and apology sincerity (F(4, 321) = 11.26, p < 0.001, η_p² = 0.12).

Conducting a simple effects analysis helped us to understand this interaction in more detail (Figure 2). When online customer service did not provide an apology, the main effect of compensation mode was not observed, as there were no significant differences in customer repurchase intention among the three monetary compensation conditions (Mnmc = 1.33, SD = 0.45 vs. Mpmc = 1.67, SD = 0.58 vs. Mimc = 1.76, SD = 0.69, all ps > 0.05). When customers received an apology, but it was casual, there was no significant difference in repurchase intention between the no-compensation group and the passive group (Mnmc = 1.92, SD = 0.72 vs. Mpmc = 2.08, SD = 0.88, p > 0.95), but it was highest in the initiative compensation group and significantly different from the first two groups (Mimc = 2.99, SD = 1.20, p < 0.001). When the customer received a sincere apology, the main effect of compensation modes was most prominent as the mean values of all three groups were statistically different, which is consistent with H2.

To better understand the influence of consumer pickiness on the interaction effect between monetary compensation and apology sincerity on repurchase intention, we employed a Hayes [66] PROCESS macro (Model 3) to conduct a moderated moderation analysis. Dummy variable coding was applied to the two categorical variables: monetary compensation (no-monetary-compensation group was encoded as the reference variable; X1 = passive, X2 = initiative) and apology sincerity (no-apology group was used to be the reference variable; W1 = casual, W2 = serious). The results (

Table 3. Bootstrap confidence intervals for the moderated moderating effect of pickiness.

Antecedent	Repurchase Intention
	Coefficient	SE	t-Value	LLCI	ULCI
Constant ***	1.34	0.16	8.61	1.04	1.65
X1	0.32	0.21	1.50	−0.10	0.74
X2	0.38	0.21	1.77	−0.04	0.80
W1 *	0.50	0.21	2.32	0.08	0.92
W2 ***	2.12	0.21	9.90	1.70	2.54
X1 × W1	−0.03	0.30	−0.11	−0.63	0.56
X1 × W2 **	0.91	0.30	3.05	0.32	1.50
X2 × W1 *	0.77	0.30	2.57	0.18	1.36
X2 × W2*	0.94	0.37	2.52	0.20	1.67
Pickiness	0.03	0.15	0.22	−0.26	0.32
X1 × pickiness	−0.19	0.23	−0.80	−0.64	0.27
X2 × pickiness	−0.22	0.22	−1.03	−0.65	0.20
W1 × pickiness	−0.36	0.19	−1.88	−0.73	0.02
W2 × pickiness	−0.20	0.24	−0.85	−0.67	0.26
X1 × W1 × pickiness	0.31	0.34	0.91	−0.36	0.97
X1 × W2 × pickiness *	0.94	0.37	2.56	0.22	1.67
X2 × W1 × pickiness	0.54	0.30	1.83	−0.04	1.13
X2 × W2 × pickiness ***	1.88	0.46	4.09	0.98	2.78
Model summary	R2 = 0.75
	F (17, 312) = 55.64 ***

Notes: SE = standard error; LLCI = lower level of 95% confidence interval, ULCI = upper level of 95% confidence interval; * p < 0.05, ** p < 0.01, *** p < 0.001.

) showed that the interaction of the passive compensation dummy variable (X1) and serious apology dummy variable (W2) with pickiness was significant (β = 0.94, SE = 0.37, t(312) = 2.56, p < 0.05), and the interaction effect was also significant for the initiative compensation dummy variable (X2) (β = 1.88, SE = 0.46, t(312) = 4.09, p < 0.001).

The significance of the interaction between monetary compensation and apology sincerity varied notably across different levels of consumer pickiness (see

Table 4. Conditional two-way interaction at different values of pickiness.

Pickiness	F	df1	df2	p
Mean − 1SD = −0.89	1.10	4	312	0.36
Mean = 0	5.53	4	312	0.00
Mean + 1SD = 0.89	10.70	4	312	0.00

). Specifically, when consumer pickiness was low, the interaction term did not significantly predict repurchase intention (F(4, 312) = 1.10, p > 0.1), indicating that the moderating effect of apology sincerity on the relationship between monetary compensation and repurchase intention was relatively weak in low-pickiness contexts. However, as pickiness increased, the interaction effect became significant (F(4, 312) = 5.53, p < 0.001), suggesting that apology sincerity played a stronger moderating role in the effect of monetary compensation. At high levels of pickiness, this interaction was most pronounced (F(4, 312) = 10.70, p < 0.001), showing that apology sincerity significantly moderated the effect of monetary compensation on repurchase intention. Overall, this pattern suggests that pickiness functions as a second moderator that amplifies the interaction between monetary compensation and apology sincerity.

Figure 3 provides a more intuitive illustration of the results shown in Table 4. Across the three pickiness groups, consumer responses were generally consistent in the no-apology and casual-apology conditions. However, in the serious-apology condition, the increase in monetary compensation elicited the strongest positive responses among highly picky consumers. These findings support H3, indicating that when a proper service recovery is offered, picky customers may exhibit even higher repurchase intentions than less picky ones.

3.1.3. Discussion

Study 1 demonstrated that monetary compensation should be initiatively offered rather than passively provided, and its effectiveness increases with either casual or serious apologies. However, our findings also indicated that when online customer service does not apologize, the differences among the groups based on the initiative of monetary compensation become negligible. This emphasizes that sincerity in apologies is more effective than a casual approach. Moreover, a casual apology only achieves its maximum effect when combined with initiative compensation. This highlights the crucial role of apologies when consumers receive products with minor defects. Furthermore, for picky consumers, the more comprehensive the service recovery measures are, the more positive their repurchase intentions become. However, less picky consumers show a relatively flat and less positive response to a serious apology from customer service regardless of the mode of compensation offered. This suggests that while a genuine apology can address the issue, it may not be sufficient to retain these customers. Our experimental results indicate that picky consumers are not always as troublesome to deal with as they are often perceived to be.

3.2. Study 2

The purpose of Study 2 was twofold. First, Study 2 aimed to enhance the robustness of H1 and H2 by replicating the results of Study 1 using a different product (a desk). Second, we tested a multiple moderation model to evaluate the moderating role of apology in the association of monetary compensation with customers’ intention to repurchase by online store reputation (H4).

3.2.1. Design and Participants

A total of 537 participants were recruited through the Wenjuanxing platform (52.89% female). Study 2 conducted a 3 (monetary compensation: initiative vs. passive vs. no) × 3 (apology: serious vs. casual vs. no) × 2 (e-store reputation: high vs. low) between-subjects design, with all participants being randomly assigned to one of the 18 groups. Table 2 displays the demographic information of the participants.

The manipulations of monetary compensation and apology sincerity were the same as in Study 1, the sole difference being the product switch to a desk with the defect described as a minor nick on the edge. The rationale for this choice was that minor defects in desks are more challenging to detect during the quality inspection process, and desks are prone to damage during transportation. The manipulation of the online store’s reputation was reflected in the written material of the scenario introduction. Participants in the high-reputation group were informed that the store from which they were purchasing the desk had a rating of 4.9 stars (out of 5 stars), while the store in the low-reputation group was rated at 3.9 stars. The overall rating was also introduced in the scenario: “E-store star ratings include three aspects: product description, service attitude, and logistics service. The highest rating is 5 stars, and ratings from 5 stars to 1 star are considered excellent, good, average, poor, and very poor in terms of the overall strength in these three aspects” (see Appendix B). As a manipulation check, participants rated the reputation on a 7-point Likert scale [67] (Cronbach’s α = 0.89). The remaining measures regarding the severity of product defects, sincerity of the apology (Cronbach’s α = 0.90), initiative of monetary compensation, and repurchase intention (Cronbach’s α = 0.93) were identical to those used in Study 1 (see Appendix C).

The results of the one-sample t-test analysis showed that participants perceived the minor nick on the desk as a defect of low severity (M = 2.68, SD = 1.10, t(536) = −6.83, p < 0.001). The high-reputation group was considered to have a better reputation than the low-reputation group through independent sample t-test analysis (Mhigh = 4.69, SD = 1.52 vs. Mlow = 3.93, SD = 1.49, t(537) = −5.90, p < 0.001). Results from a one-way ANOVA revealed that the initiative of monetary compensation among the three groups (initiative, passive, and no-compensation group) ranked from highest to lowest as follows: (Mimc = 5.22, SD = 1.56 vs. Mpmc = 4.72, SD = 1.67 vs. Mnmc = 2.85, SD = 1.55, F(2, 534) = 109.69, all ps < 0.01). The subjects in the serious-apology group were also perceived by participants to be more genuine compared to subjects in the casual- and no-apology group (Msap = 4.79, SD = 1.32 vs. Mcap = 4.39, SD = 1.63 vs. Mnap = 3.39, SD = 1.46, F(2, 534) = 42.80, all ps < 0.05). Therefore, all manipulations were successful.

3.2.2. Results

The results of the two-way ANOVA are consistent with Study 1, showing significant main effects of both the initiative of monetary compensation (F(2, 528) = 14.83, p < 0.001, η_p² = 0.05) and the sincerity of apologies on repurchase intention (F(2, 528) = 35.70, p < 0.001, η_p² = 0.12). Customers in the initiative compensation group (Mimc = 4.54, SD = 1.59) exhibited significantly higher repurchase intentions than those in both the passive (Mpmc = 4.17, SD = 1.55, p < 0.05) and no-compensation groups (Mnmc = 3.71, SD = 1.54, p < 0.001). There was a significant interaction effect between the sincerity of apologies and monetary compensation on repurchase intention (F(4, 528) = 4.83, p < 0.001, η_p² = 0.04).

A simple effects analysis revealed that the main effect of the monetary compensation mode on repurchase intention disappeared in the no-apology group (Mnmc = 3.44, SD = 1.67 vs. Mpmc = 3.47, SD = 1.33 vs. Mimc = 3.48, SD = 1.47, all ps > 0.05), but it was prominent in the serious-apology group (Mnmc = 4.00, SD = 1.32 vs. Mpmc = 4.64, SD = 1.40 vs. Mimc = 5.60, SD = 0.99, all ps < 0.05). In cases of casual apologies, there were significant differences between the no-compensation (Mnmc = 3.69, SD = 1.60) and passive (Mpmc = 4.42, SD = 1.66, p < 0.05) as well as between no-compensation and initiative groups (Mimc = 4.55, SD = 1.45, p < 0.01). However a statistically significant difference was not observed between the passive compensation and initiative compensation groups (Mpmc = 4.42, SD = 1.66 vs. Mimc = 4.55, SD = 1.45, p > 0.05). This suggests that apologies play a moderating role in the relationship between monetary compensation and repurchase intention (Figure 4).

We adopted the same analytical approach as in Study 1 to test H4 (

Table 5. Bootstrap confidence intervals for the moderated moderating effect of reputation.

Antecedent	Repurchase Intention
	Coefficient	SE	t-Value	LLCI	ULCI
Constant ***	3.57	0.61	5.88	2.38	4.76
X1	0.20	0.84	0.24	−1.45	1.85
X2	0.08	0.84	0.10	−1.57	1.73
W1	−1.16	0.85	−1.36	−2.82	0.51
W2	0.98	0.84	1.16	−0.68	2.63
X1 × W1	1.99	1.20	1.67	−0.36	4.34
X1 × W2	−0.53	1.18	−0.45	−2.84	1.79
X2 × W1 **	3.46	1.18	2.92	1.13	5.78
X2 × W2 *	0.47	1.18	0.39	−1.86	2.79
Reputation	−0.09	0.38	−0.22	−0.84	0.66
X1 × reputation	−0.12	0.53	−0.22	−1.16	0.93
X2 × reputation	−0.03	0.53	−0.06	−1.07	1.01
W1 × reputation	0.94	0.54	1.75	−0.11	1.99
W2 × reputation	−0.28	0.53	−0.52	−1.32	0.77
X1 × W1 × reputation	−0.86	0.75	−1.14	−2.34	0.62
X1 × W2 × reputation	0.76	0.75	1.02	−0.71	2.23
X2 × W1 × reputation *	−1.76	0.75	−2.34	−3.23	−0.29
X2 × W2 × reputation	0.73	0.75	0.98	−0.74	2.20
Model summary	R2 = 0.21
	F (17, 519) = 7.96 ***

Notes: SE = standard error; LLCI = lower level of 95% confidence interval, ULCI = upper level of 95% confidence interval; * p < 0.05, ** p < 0.01, *** p < 0.001.

and

Table 6. Conditional two-way interaction at different values of reputation.

Reputation	F	df1	df2	p
Low	3.22	4	519	0.01
High	4.73	4	519	0.00

), and a significant interaction effect among monetary compensation initiation, apology sincerity, and reputation was found (F(4, 519) = 3.06, p < 0.05). Specifically (Figure 5), in the no-apology group, regardless of the store’s high (Mimc = 3.42, SD = 1.53 vs. Mpmc = 3.36, SD = 1.30 vs. Mnmc = 3.40, SD = 1.58, all ps > 0.05) or low reputation (Mimc = 3.53, SD = 1.44 vs. Mpmc = 3.56, SD = 1.38 vs. Mnmc = 3.48, SD = 1.78, all ps > 0.05), customers’ repurchase intentions among the three monetary compensation groups were not statistically significant.

However, when the online customer service apologized casually, there was a significant difference in consumers’ repurchase intentions between compensation versus no compensation under the low-reputation condition (Mimc = 5.02, SD = 1.37 vs. Mnmc = 3.27, SD = 1.59, p < 0.001; Mpmc = 4.48, SD = 1.87 vs. Mnmc = 3.27, SD = 1.59, p < 0.01). It shows that a casual apology can enhance the effect of monetary compensation modes in the low-reputation group, but this was not the case under the high-reputation condition (Mimc = 4.08, SD = 1.48 vs. Mpmc = 4.36, SD = 1.49 vs. Mnmc = 4.12, SD = 1.53, all ps > 0.05). In the case of high-reputation e-stores, significant differences in the three groups of monetary compensation initiation are observed only when customers receive a serious apology (Mimc = 5.78, SD = 1.11 vs. Mpmc = 4.48, SD = 1.51 vs. Mnmc = 3.82, SD = 1.33, all ps < 0.01). However, in the case of low-reputation e-stores, the initiative and passive monetary compensation can bring higher repurchase intention even in the casual-apology mode.

3.2.3. Discussion

Study 2 replicated the findings of Study 1 concerning the impact of apology sincerity and monetary compensation on repurchase intentions. In terms of store reputation, interestingly, in cases of product defects with compensation, a good reputation paradoxically acts as a disadvantage. For high-reputation e-stores, the initiation of monetary compensation exerts a positive effect on consumers’ intentions to repurchase but only in the context of a serious apology. However, for low-reputation e-stores, a casual apology accompanied by compensation results in higher consumer repurchase intentions than no compensation, but not for high-reputation e-stores.

4. General Discussion

This study compares the effectiveness of various service recovery strategies for products with minor defects and investigates in depth the interplay between these strategies at the varying levels of consumer pickiness, as well as the reputation of e-stores on consumers’ repurchase intentions. The research results have produced some theoretical and practical implications, along with directions for future research.

4.1. Theoretical Contributions

First, this study augments the literature on service recovery for minor-defect products. Previous scholars across different countries have provided varying answers to which service recovery measures should be applied in different online service failure scenarios [2,11,20]. Moreover, past service recovery approaches have been limited to the binary presence or absence of apologies and compensation [10,12,21,22]. To fill these gaps, through two scenario-based experiments, we explored how the interaction between the sincerity of apologies and the initiation of monetary compensation can bolster consumer confidence in revisiting the store. The results indicate that in scenarios where consumers receive products with minor defects from online shopping, passive provision of monetary compensation is less effective than initiative provision. Additionally, consumers’ repurchase intentions are lower when receiving a casual apology compared to a serious one, and an at least casual apology needs to be combined with initiative compensation to be effective. More importantly, we propose that compared to a casual apology, a serious apology enhances the positive effect of compensation initiative on consumer repurchase intentions, but the absence of an apology nullifies the effectiveness of initiative compensation. This study not only extends online service recovery strategies to different application scenarios but also offers a detailed and comprehensive analysis and predictive insights into two pivotal remedial actions: apologies and compensation.

Second, we introduced the concept of consumer pickiness, a significant personality trait, to explore variations in the moderating effects of apologies across different consumer groups. Prior research often neglected the segment of picky consumers, frequently conflating them with maximizers and disproportionately emphasizing their negative traits [26,27]. In our study, picky consumers demonstrated unexpectedly higher intentions to repurchase when they received a sincere apology and initiative compensation from online customer service, in stark contrast to those with lesser degrees of pickiness. The core characteristic of picky consumers is not an absolute intolerance of defects but rather a heightened sensitivity to quality and service, along with more clearly defined expectations. Instead of immediately abandoning a brand after a failure, they are more likely to reassess its trustworthiness based on how the recovery is handled. This finding provides further support and insight into the distinct nature of pickiness as a consumer trait. Additionally, the moderating effect of apology is more pronounced when customers are picky. In other words, when customers are not picky, apology does not sufficiently enhance the effect of the mode of monetary compensation. However, for picky customers, the apology has a stronger effect on enhancing the main effect of the modes of monetary compensation, especially in the serious apology condition.

Third, aiming to establish a detailed guide to the framework for service recovery strategies specifically for minor product defects, our study has enriched the existing literature by emphasizing the critical role of store reputation in the context of service recovery. Prior research has maintained that a high reputation should not be used as an excuse for customers to accept low-effort recovery strategies (e.g., offering only monetary compensation) [21]. Our research results not only support the notion but also go a step further. The results of this research show that especially in the case of high-reputation stores, even initiative compensation alone does not achieve a better service recovery than passive or no compensation. However, when the monetary compensation is accompanied by a serious apology, the effect of monetary compensation modes is strengthened in that initiative monetary compensation shows higher service recovery compared with other modes. High-reputation e-stores induce preconceived trust and favorability among consumers. Thus, when a service failure occurs and fails to meet their expectations, a serious apology becomes particularly important in the service recovery process for the e-stores with high reputation. However, low-reputation e-stores inherently do not set high expectations among consumers and may not require a profound apology to restore confidence.

Fourth, this study adopts a prospective perspective to examine service failure and recovery among Chinese consumers in Asia, an important topic in the global research agenda. Previous studies have suggested that compensation is more suitable for countries with an individualistic cultural orientation (e.g., the United States) and that apology tends to be more effective than compensation in Asian countries [4]. However, our findings suggest that compensation retains substantial value in service recovery in the Asian market. Specifically, initiative monetary compensation exerts a stronger influence on service recovery outcomes than passive compensation. Moreover, we find that apologies perceived as highly sincere significantly enhance the effectiveness of monetary compensation, whereas casual apologies offer little additional benefit. This study fills a regional empirical gap and provides practical guidance for brand managers to develop globally consistent yet culturally adaptive service recovery strategies in the Asian market.

4.2. Practical Implications

E-retailers should avoid neglecting an apology when offering monetary compensation, as this oversight can easily prove counterproductive. Furthermore, even when providing compensation, it is crucial to maintain a proactive stance. Our results indicate that waiting for customers to demand compensation is less effective than retailers taking the initiative. Therefore, when monetary compensation by e-commerce retailers is inevitable, they should swiftly seize the opportunity to take proactive action and assertively maximize the efficacy of their service recovery efforts.

Our results suggest that e-retailers should be pleased when encountering picky consumers seeking customer service. In the long run, the benefits of winning back such consumers far outweigh the increased costs associated with compensation. We recommend e-retailers not to skimp on proactivity and sincerity when dealing with picky customers. However, it is also important to recognize that overly generous service recovery strategies may carry certain risks. If compensation or apologies appear too lenient, they may encourage opportunistic customer behaviors, such as deliberately exaggerating problems to gain additional benefits. While less picky customers are generally considered easier to handle, this does not imply that they will provide additional opportunities for store visits. Our study reveals that a sincere apology resonates positively with less picky customers, but offering compensation does not notably enhance this effect. Therefore, e-retailers should consider employing sincere verbal apologies with less picky customers, thus avoiding the unnecessary expenditure of resources.

Overall, firms should not rely solely on verbal apologies after service failures but should adopt a combined strategy of apology and compensation. For high-reputation stores, customers expect high standards, making perfunctory recovery efforts ineffective. Only a serious apology accompanied by initiative monetary compensation can sustain trust and repurchase intention, reflecting the responsibility and professionalism consistent with their reputation. For low-reputation stores, monetary compensation plays a central role in rebuilding consumer trust. Without the credibility of reputation, compensation functions as a costly signal that helps break entrenched negative perceptions. Even a casual apology, when paired with compensation, can significantly enhance repurchase intention. This suggests that in low-reputation stores, the effectiveness of recovery depends less on the perfection of the apology and more on whether the compensation conveys genuine sincerity. If the response exceeds customers’ low expectations, it can potentially foster improved word of mouth and trust restoration.

4.3. Limitations and Future Research

In the context of service recovery, this research shows that the service recovery strategies using apologies and monetary compensation seem to be more effective for picky consumers compared to those for less picky consumers. This phenomenon might stem from the fact that less picky consumers do not exert as much effort in their shopping endeavors as their picky counterparts, making them less perturbed by a single unfavorable experience and more inclined to simply choose not to revisit. Future studies could investigate whether alternative service recovery strategies, such as providing explanations for the failure or guarantees against future occurrences, would be more impactful for this consumer segment.

Moreover, our study focused solely on one dependent variable, repurchase intention. However, considering the ubiquity of social media, unsuccessful service recovery for minor product defects may prompt consumers to vent their frustrations on social networks. Prior research has demonstrated that negative word of mouth can significantly affect an e-retailer’s image among potential customers [68]. Therefore, it is imperative to investigate which service recovery measures might not only maintain repurchase intentions but also potentially prevent unwarranted negative word of mouth.

The experiments in this study were conducted with Chinese participants, which may introduce certain cultural limitations. Individuals from different cultural backgrounds may vary in their perceptions of apology norms, sincerity, and compensation expectations. Therefore, future research is encouraged to replicate and extend these findings in cross-cultural contexts to enhance the external validity and generalizability of the results. In addition, this study employed online scenario-based experiments, which may not fully capture the judgments and behaviors individuals exhibit in real purchase situations. Future research could adopt more ecologically valid methods, such as behavioral experiments or field studies, to reduce potential response biases arising from hypothetical scenarios.