Introduction

Managers can enhance proactive customer service behavior (hereinafter, PCSB), which is characterized by self-initiated and anticipatory actions where service providers proactively assume responsibility for addressing service-related issues to enhance customer satisfaction (Raub & Liao, 2012). PCSB improvement can be achieved through the cultivation of a conducive work environment—for example, by promoting employee initiative (Hong et al., 2016), fostering affective commitment (Rank et al., 2007), and implementing transformational leadership (Den Hartog & Belschak, 2012). Although feedback has long been considered a developmental and learning tool that motivates employees to align their efforts and goals with performance standards (Ashford, 1986; Chan & Lam, 2011; De Stobbeleir et al., 2011), its role in promoting PCSB has received limited attention. This gap has important implications for managers. For example, it is unclear whether a manager trying to encourage employees to seek and act on opportunities to enhance PCSB should provide positive or negative (or mixed) performance feedback. We offer theoretical insights into this question and practical, actionable guidance for managers.

We argue that this gap in understanding cannot be satisfactorily addressed without a comprehensive synthesis of the feedback, PCSB, and goal-setting literature. Relying solely on feedback to enhance PCSB may prove inadequate. While considerable research has explored the connection between feedback and goal setting (Locke & Latham, 2012; Renn & Fedor, 2001), limited attention has been given to the impact of feedback on PCSB. Notably, because PCSB is widely acknowledged as a goal-driven process that requires deliberate and voluntary actions (Parker et al., 2010), we propose that goal-setting behavior is the crucial link between feedback and PCSB. Despite the significance of feedback, goal setting, and PCSB in managerial discourse, these topics have predominantly been studied in isolation in the marketing literature. Surprisingly, there is a paucity of research examining the intersection of these three domains and the mechanisms that connect them, resulting in a limited understanding of the influence of feedback on PCSB.

To bridge this research gap, we explore the underlying mechanisms that elucidate how positive and negative feedback impacts the feedback-driven goal-setting process while investigating potential differences between the two types of feedback. Our research delineates how managers can utilize feedback as a tool for enhancing customer service. Our findings reveal that managers should promote a sense of utility when delivering positive feedback and a sense of accountability when issuing negative feedback to ensure that it effectively contributes to improving PCSB. In addition, our findings inform managers that goal setting, informed by feedback, only boosts PCSB in a work environment that supports self-initiation and highlights the importance of an organizational climate that fosters initiative.

While it is known that employees receive positive, negative, and mixed feedback from managers, the process by which this information affects feedback-based goal setting remains ambiguous. Research suggests that positive feedback operates differently from negative feedback; thus, more attention must be given to the nuanced differences in how they may affect feedback-based goal setting (Fishbach & Finkelstein, 2012; Förster et al., 2001; Idson & Higgins, 2000). To the best of our knowledge, prior research has not explored the specific mechanisms by which positive and negative feedback influence feedback-based goal setting and, subsequently, PCSB.

Given this background and to address our research question, we introduce the Motivational driver–Goal setting–Goal striving–Goal attainment, or MG3, model. Figure 1 shows these different elements of the model, which correspond to the constructs depicted in our conceptual framework. This model illustrates how motivational drivers, such as positive and negative feedback, influence goal attainment (i.e., customer service performance) through goal setting (i.e., feedback-based goal setting) and goal striving (i.e., PCSB). The MG3 model is a comprehensive goal process model that builds on the intersection of feedback giving, goal setting/striving, and proactive service behavior literature (for details, see Web Appendix A).

Fig. 1
figure 1

Model (Study 1). Notes: (1) T = time. (2) Extra-role behavior is included in the model for assessing the robustness of the proposed model (i.e., alternative mediator in the relationship between feedback-based goal setting and customer service performance). (3) The dashed line connecting initiative climate and feedback-based goal setting represents a relationship that is not formally hypothesized but is estimated. (4) The dashed line connecting proactive customer service behavior and customer service performance represents a relationship that is not formally hypothesized but is examined to assess the nomological validity of the model

Full size image

This research makes four primary contributions. First, we test the motivational driver–goal setting–goal striving part of the MG3 model by examining how positive and negative feedback influence PCSB through feedback-based goal setting. The results show that feedback-based goal setting is the conduit through which positive feedback affects PCSB, highlighting the importance of using positive feedback to set goals if employees want to engage in PCSB.

Second, we demonstrate the different underlying mechanisms by which positive and negative feedback influence feedback-based goal setting. Studies have not fully expounded on the underlying processes steering this relationship. As the literature suggests, while positive feedback strengthens employees’ goal persistence and commitment, negative feedback reveals a lack of goal progress (a discrepancy between the current and the desirable end state) and the need to make adjustments (Fishbach et al., 2010; Fishbach & Finkelstein, 2012). We extend the literature by revealing the paths from positive and negative feedback to feedback-based goal setting. Our results support different mediation pathways whereby positive feedback affects feedback-based goal setting through feedback utility while negative feedback does so via feedback accountability (Linderbaum & Levy, 2010).

Third, we find that feedback-based goal setting has a positive impact on PCSB only when initiative climate is high, but not when it is low. This suggests that managers cannot expect to evoke PCSB despite feedback-based goal setting if initiative climate is low. This draws attention to the boundary conditions of when feedback-based goal setting can effectively encourage PCSB, adding nuance to the literature.

Fourth, the MG3 model includes extra-role behavior (ERB) as an alternative goal-striving behavior in addition to PCSB, considering that not all ERBs are PCSBs and not all PCSBs are ERBs. For example, some PCSBs may be part of an employee’s formal job description, while ERBs may be a reactive response to customer complaints or requests (Crant, 2000; Parker & Collins, 2010). Empirically, few studies have examined ERB and PCSB in the same model, raising concerns about alternative or omitted goal-striving behaviors. We find that feedback-based goal setting has no effect on ERB (only on PCSB) and that PCSB positively affects customer service performance even after controlling for ERB. By including ERB and PCSB concurrently in our model, we increase the validity and robustness of our results.

We organize the remaining sections of this paper as follows. Next, we discuss the MG3 model, which provides the foundation for our hypotheses development. Then, we test our hypotheses through three studies. Study 1 uses multilevel, multi-respondent, and time-lagged survey data from the hospitality industry to test the entire MG3 model. Study 2 is a controlled experiment in which we manipulate the valence (positive vs. negative) of feedback to test the mediating roles of feedback utility and accountability between positive/negative feedback and feedback-based goal setting. In Study 3, we replicate and extend Study 2 by examining not only positive and negative feedback but also mixed feedbackFootnote 1 (50–50 mix of positive and negative). We conclude by discussing the theoretical and managerial implications of our work, its limitations, and suggestions for further research.

Theoretical framework: The MG3 model

Goals setting based on feedback

Goal-setting theory is a motivational theory that encourages goal-directed behaviors to facilitate goal striving and attainment by promoting purposeful self-regulation (Latham & Locke, 1991; Locke & Latham, 1990). Goal striving involves behaviors such as persistence, commitment, planning, and purpose that help people convert their goals into desired outcomes. Setting goals enables people to concentrate their resources (e.g., time, effort) on goal-related activities that facilitate the achievement their objectives. While several factors can influence goal setting, such as goal desirability, feasibility, and relevancy (Bagozzi & Dholakia, 1999), we focus on feedback as a motivational driver that can reinforce goal commitment via positive feedback and motivate goal progression via negative feedback (by signaling a discrepancy between one’s current performance level and one’s target) (Fishbach & Finkelstein, 2012). Feedback provides instrumental and diagnostic information to regulate goal-setting behavior, and feedback-based goal setting plays a vital role in the MG3 model (Renn & Fedor, 2001).

Goal-setting theory asserts that feedback should be utilized to set goals or facilitate performance improvement (Locke & Latham, 1990). Feedback should lead to the setting of and commitment to specific, challenging goals to increase motivation and improve performance (Latham & Locke, 1991). Thus, an essential aspect of goal-setting theory is that a manager’s feedback empowers employees to set performance goals. Feedback and goal setting are closely connected, as noted by Locke and Latham (2012, p. 626), who state, “Goals and feedback together work better than either one alone.” Reflecting this interdependent relationship, prior studies have explored the role of feedback-based goal setting, which refers to employees establishing and updating improvement-related objectives based on performance feedback provided by supervisors (Renn & Fedor, 2001). This construct integrates feedback with goal setting and underscores that employee performance is significantly enhanced when both feedback and goal setting are utilized in tandem, rather than one at the expense of the other (Ashford & De Stobbeleir, 2013).

The MG3 model (see Fig. 1) is integrates goal-setting theory with the feedback and proactive service behavior literature (Ashford & De Stobbeleir, 2013; Kluger & DeNisi, 1996). The model is comprehensive because it encompasses the full range of motivational factors that contribute to goal setting, such as positive and negative feedback, as well as striving for goals. In our context, this applies to improving PCSB through feedback-based goal setting (Kluger & DeNisi, 1996; Neubert, 1998). Goal setting acts as an intermediary between feedback and PCSB. However, feedback must be incorporated into goal setting. As Ashford and De Stobbeleir (2013, p. 51) emphasize, “Performance is improved most when both feedback and goals are present.”

Feedback utility and feedback accountability

The use of negative feedback has broad implications ranging from eliciting emotions such as shame (Xing et al., 2021) to affecting usage behavior (Bogard et al., 2020) and creativity (Kim & Kim, 2020). Recent literature on feedback and goal setting affirms that goal attainment involves a dedicated and committed process that entails goal setting and striving (Giessner et al., 2020; Silverman et al., 2023). Yet the exact mechanism of how negative feedback differs from positive feedback in influencing feedback-based goal setting remains unclear. The proposed model unpacks the relationship between positive and negative feedback and feedback-based goal setting by positioning feedback utility and accountability as mediators. By positive (negative) feedback, we refer to the information managers provide to employees when an employee’s performance on key customer-related dimensions such as service quality, responsiveness, product and service knowledge, social interactions, and added value exceeds (falls short of) a standard/expectation or goal.

As we explain in the “Hypotheses development” section, the indirect path from positive feedback to feedback-based goal setting differs from the indirect path from negative feedback to feedback-based goal setting. This difference is mainly due to the different motivations these two types of feedback provide. We argue that feedback utility and accountability capture these differences as mediators in the model (Fishbach et al., 2010; Fishbach & Finkelstein, 2012; Linderbaum & Levy, 2010).

Feedback utility refers to employees’ belief that feedback will help them achieve their goals (Linderbaum & Levy, 2010). This concept is rooted in the “belief in the value of feedback” (London & Smither, 2002, p. 83), which suggests that feedback can produce other desirable outcomes. Drawing on Vroom’s (1964) Valence Instrumentality Expectancy Theory, feedback utility reflects the motivation to use feedback based on its instrumentality (i.e., the ability of feedback to lead to goal attainment) and expectancy (i.e., the likelihood that feedback will lead to goal attainment). Feedback utility considers feedback from a prescriptive perspective, assuming that it will result in desirable outcomes.

In contrast, feedback accountability is concerned with employees’ tendency to react to and act on feedback (Linderbaum & Levy, 2010) and draws on the concept of “feeling accountable to act on feedback” (London & Smither, 2002, p. 83). Accountability captures a sense of obligation on the part of employees to use feedback from their managers. Feedback accountability considers feedback from a normative perspective, emphasizing a sense of responsibility to use feedback for goal setting.

Hypotheses development

Recall that the MG3 model is an integrated yet comprehensive conceptual framework that synthesizes goal theory with the feedback and proactive behavior literature. This model asserts that manager feedback is a motivational driver, regardless of feedback valence, with positive and negative feedback motivating employees in different ways: utility for positive feedback and accountability for negative feedback. The model also proposes that employees use feedback as an important source of input to engage in self-regulatory behaviors such as feedback-based goal setting and PCSB. Self-regulatory behaviors refer to processes by which employees control, manage, and direct their thoughts, emotions, and actions to achieve specific objectives or adhere to standards of behavior required by managers. In this self-regulatory process, we posit that a supportive work environment, such as initiative climate, will act as a facilitator by providing support that aligns with what is needed to deliver PCSB successfully. Ultimately, when employees engage in a self-regulatory process, their performance is likely to be elevated because employees can adjust and modify their behavior to serve customers better. The core logic of the hypotheses examined in this research is grounded in the MG3 model. We present our reasoning in the subsequent sections.

The mediating effect of feedback-based goal setting

Feedback can motivate employees for different reasons (Camacho et al., 2019; Finkelstein & Fishbach, 2012). However, regardless of valence (i.e., positive or negative), feedback-based goal setting increases when employees understand that feedback is beneficial and “the right thing to do” when setting goals. Positive feedback reinforces and reassures employees that they are on the right track, incentivizing them to continue with their current approach when setting goals (Finkelstein & Fishbach, 2012). Positive feedback tends to boost motivation and self-confidence. It reinforces individuals’ beliefs that what they are currently doing is working and that goal setting will allow them to stay on track (Fishbach et al., 2010). For example, if an employee receives positive feedback about his or her customer empathy, that employee will likely believe that the current approach is effective and continue to show empathy for the customer when setting goals.

In contrast, negative feedback highlights a performance gap between the employee’s current service performance and how the manager expects the employe to perform, thus providing valuable information to track progress. Negative feedback can also motivate change and elicit reflection by highlighting areas of weakness or underperformance (Finkelstein & Fishbach, 2012). By identifying areas for adjustment, negative feedback can provide a clear direction for goal setting in areas where improvements are needed most.

A potential counterargument can be considered, as studies have demonstrated that negative feedback can hinder employee creativity. This occurs through meta processes by diminishing one’s self-concept, ego, and social image (Kim & Kim, 2020). Given our focus on goal setting over creativity and informed by literature suggesting that negative feedback serves as a prompt for progress evaluation, we propose that negative feedback will have a constructive impact on feedback-based goal setting.

Furthermore, PCSB is considered a goal-striving behavior, as proactive behavior requires purposeful and voluntary actions toward improving customer service (Rank et al., 2007). We propose that after employees set customer service improvement goals based on feedback, they will take self-regulatory actions to strive toward such goals. This involves an active and forward-thinking approach to customer service rather than a reactive and passive one. Feedback-based goal setting leads to PCSB because setting goals based on feedback provides direction and allows employees to allocate and redistribute their resources in a deliberate, targeted, and disciplined manner (Parker & Collins, 2010). Specifically, feedback-based goal setting provides information about where to focus more or less effort and time to deliver PCSB.

Proactive behavior research has shown that cognitive and motivational factors (e.g., role breadth efficacy, flexible orientation) mediate the relationship between proactive personality and proactive behavior (Parker et al., 2006). However, the role of goals in the proactive behavior process has yet to be fully explored empirically, although goal-setting theory can inform the understanding of proactive behaviors (Parker et al., 2010). As Crant (2000, p. 454) argues, “goals may mediate the relationship between individual differences and contextual factors and proactive behaviors. Context and personality may yield specific goals best achieved through the exhibition of proactive behavior.”

Integrating the foregoing arguments, we posit that feedback-based goal setting mediates the relationship between positive/negative feedback and PCSB. Feedback-based goal setting can be a transitional stage that allows employees to translate positive/negative feedback into PCSB. Without feedback-based goal setting, positive/negative feedback alone may be insufficient to motivate employees to engage in PCSB, as employees may not act on the feedback (Neubert, 1998). Only when feedback is integrated into the act of goal setting can it be implemented and unleash its potential to elicit behavioral change. Furthermore, the mediating role of feedback-based goal setting aligns with growing calls in the literature to include goals to explain the proactive behavior process (Crant, 2000; Parker et al., 2010). Therefore, we argue that feedback-based goal setting serves as a mediator by directing how time and effort are allocated in a more deliberate, targeted, and purposeful manner.

H1a

Feedback-based goal setting mediates the effect of positive feedback on PCSB.

H1b

Feedback-based goal setting mediates the effect of negative feedback on PCSB.

Moderating effect of initiative climate

Initiative climate refers to employees’ perception that management expects and encourages them to challenge themselves, take charge, and persist in addressing service-related problems (Raub & Liao, 2012; Sok et al., 2021). Such a climate stimulates proactive behavior by fostering employees’ initiative to solve service-related issues without waiting for management directives (Parker et al., 2010). In a high-initiative climate, employees are encouraged to take ownership, anticipate issues, and act proactively instead of reacting to problems (Parker & Collins, 2010; Sok et al., 2021). We conceptualize initiative climate at the group level, representing the shared perception among employees that management expects, supports, and rewards proactive behavior in addressing service-related issues (Hong et al., 2016).

We suggest a cross-level interaction effect between feedback-based goal setting and initiative climate on PCSB. PCSB involves “taking initiative in improving current circumstances or creating new ones; it involves challenging the status quo rather than passively adapting to present conditions” (Crant, 2000, p. 436). PCSB entails service providers proactively assuming responsibility for service-related matters and exhibiting self-initiated and anticipatory behaviors to enhance customer service (Raub & Liao, 2012). Unlike reactive behavior that responds to customers’ requests, complaints, or failures, PCSB adopts an active orientation and aims to prevent issues from happening in the first place (Challagalla et al., 2009). Therefore, an initiative climate aligns with and reinforces proactive behavior, strengthening the effect of feedback-based goal setting on PCSB. Proactive behavior, including PCSB, involves uncertainty and risk of failure because it challenges the status quo (Parker et al., 2006). Consequently, even when employees set feedback-based goals, they may hesitate to follow through and engage in PCSB.

Nevertheless, because an initiative climate encourages and supports employees to initiate action, be proactive, and display change-oriented behavior, it can facilitate the transition from feedback-based goal setting to PCSB. We predict that the effect of feedback-based goal setting on PCSB will be amplified by a high initiative climate that can act as a buffer, absorbing some of the risks and perceived costs (e.g., reputation loss) associated with PCSB. Based on the proactive motivation framework (Parker et al., 2010), initiative climate can provide a conducive environment for feedback-based goal setting to have a positive effect on PCSB because employees perceive a “reason to” and are “energized to” engage in proactive behavior despite the uncertainty and risk associated with it.

However, if initiative climate is low, PCSB may be compromised, even if feedback-based goal setting is in place. According to social information processing theory, this is because employees will perceive that self-initiated and anticipatory actions are neither supported nor valued (Salancik & Pfeffer, 1978). When employees feel that change-oriented behaviors that entail risk and potential resistance are not encouraged, the impact of feedback-based goal setting on PCSB may be limited (Baer & Frese, 2003; Raub & Liao, 2012). Therefore, we predict that feedback-based goal setting will lead to more PCSB when initiative climate is high (vs. low).

H2

Initiative climate moderates the relationship between feedback-based goal setting and PCSB such that the relationship is stronger when initiative climate is high (vs. low).

It is worth noting that even though we do not formally hypothesize a relationship between PCSB and customer service performance, we test this relationship for the model’s nomological validity.

Overview of studies

We conducted three studies to examine the proposed model. Study 1 tests H1 and H2. We used time-lagged, multilevel, and multi-respondent data form employees and managers to address causality and common method bias concerns. Study 2 is a controlled experiment in which we manipulated feedback valence (positive vs. negative) and explored the role of feedback utility and accountability as the underlying mechanisms in feedback-based goal setting (H3). Study 3 attempts to replicate the findings from Study 2. For Study 3, we expanded the scope of our investigation by incorporating mixed feedback (i.e., a 50–50 mix of positive and negative feedback) alongside positive and negative feedback. We also used a different service context.

Study 1

Research context, sample, and data collection

We conducted Study 1 at hotels in Istanbul with the assistance of a market research company. Hotels in Istanbul are divided into three groups in terms of ownership (or management): independent hotels, local chain hotels, and global chain hotels. There are 2640 hotels and 55 hotel chains operating in Istanbul, 2509 of which are independent hotels. Most of the chain hotels in Istanbul are owned (or managed) by global chains, and these constitute approximately 22% of the room capacity. There are 194 five-star hotels in Istanbul with a capacity of approximately 29,000 rooms, most of which are operated by global hotel chains (https://www.hotelchains.com/en/turkey/istanbul/; Tourism Market Overview, 2020). In our study, we targeted five-star hotels owned (or managed) by global chains. Before collecting data, we contacted the human resources departments of 59 five-star hotels to provide information about the scope and purpose of the study and to obtain the necessary permissions. At the end of this initial process, 22 five-star hotels agreed to support our study. We provide detailed information on the hotels’ features and an overview of their performance evaluation systems in Web Appendix B.

We targeted service employees and managers working at the front desk and in the food-and-beverage units of these hotels. Because service employees directly engage with customers, they are an ideal sample group for our study. We collected data from service employees at three different times via online survey links. We informed the respondents that the hotel management endorsed the study, that their participation was entirely voluntary, and that they could withdraw at any stage. We assured them that the individual data collected would not be disclosed. We also informed the participants that they were consenting to data collection by completing the survey. Service employees responded to the surveys during work hours.

At Time 1, we sent a survey to 678 service employees, who answered questions about demographics and responded to the scales related to control variables and managers’ positive and negative feedback. One month later (Time 2), we sent the second survey to 593 service employees who had returned the first survey, and they responded to the scale on feedback-based goal setting. One month after that (Time 3), we sent a survey to 526 service employees who responded to the Time 2 survey to measure PCSB. At this stage, we also sent questionnaires to 44 managers to evaluate the service performance of the service employees working under their supervision. After matching the returned and usable surveys across three waves, we obtained 466 service employee surveys (an effective response rate of 69.8%). We received full survey participation from managers, and all surveys were usable. As a result, the dataset consisted of 466 service employee–manager dyads. Because we collected data from service employees at the front desk and food-and-beverage units, we grouped the service employees under two units in each hotel for future analyses (i.e., 44 work units in 22 hotels). Of the service employees, 69% were male, the average age was 31 years (SD = 4.4), 71% were college graduates, the average work experience was 6.1 years (SD = 4.5), and the average hotel experience was 3.8 years (SD = 3.1). Of the managers, 61% were male, the average age was 42 years (SD = 13.4), 89% were college graduates, the average work experience was 8.9 years (SD = 9.9), and the average hotel experience was 8.6 years (SD = 4.1).

Measures

We developed scales specifically for this study to measure managers’ positive and negative feedback. We measured all other variables using scales available in the literature. Because we conducted the survey in Turkish, we used the translation and back-translation method to translate all scales into Turkish (e.g., Brislin, 1986). In addition to the core variables, we used five control variables (i.e., customer orientation, leader–member exchange, organizational identification, feedback accuracy, and work experience). We included ERB in the model as an additional mediating variable to evaluate the robustness of the proposed model. We also controlled for the effect of initiative climate when estimating feedback-based goal setting, PCSB, and customer service performance. We provide the details about the measures and scale development in Web Appendix C.

Analytical approach

The analytical approach involves two steps. First, we conducted confirmatory factor analyses to assess the reliability, validity, and unidimensionality of the measures. Second, we estimated the model by (1) accounting for observed and unobserved heterogeneity, (2) correcting for endogeneity bias, and (3) performing three-level random effects (intercepts and slopes) modeling with Bayesian estimation to test mediation and moderation effects (see Web Appendix C).

Results

Initial findings

Before testing the hypotheses, we estimated the main effects–only model to obtain preliminary findings regarding the direction and nature of the relationships shown in Fig. 1. The results suggest a significant relationship between positive feedback and feedback-based goal setting. However, the relationship between negative feedback and feedback-based goal setting is not significant. The relationship between feedback-based goal setting and PCSB is significant, as is the relationship between PCSB and customer service performance (see Web Appendix C, Table WC.5).

Mediated effects

To test H1 and H2, we estimated the mediation model (Table 1) and computed the direct, indirect, and total effects of positive and negative feedback on PCSB (see Web Appendix C, Table WC.6).

Table 1 Results of the mediation model (Study 1)
Full size table

Table 1 reveals a significant relationship between positive feedback and feedback-based goal setting (γ = 0.146, 95% confidence interval [CI] [0.082, 0.210]) and between feedback-based goal setting and PCSB (γ = 0.308, 95% CI [0.145, 0.474]). Positive feedback has a direct effect on PCSB (γ = 0.123, 95% CI [0.071, 0.175]), and its indirect effect through feedback-based goal setting is also significant (γ = 0.044, 95% CI [0.017, 0.080]). Feedback-based goal setting explains 26.9% of the positive feedback–PCSB relationship.Footnote 2 These results support H1a, such that feedback-based goal setting partially mediates the relationship between positive feedback and PCSB.

Table 1 indicates that negative feedback has no significant effect on feedback-based goal setting (γ = 0.034, 95% CI [–0.017, 0.084]). Additionally, neither its direct effect (γ = 0.037, 95% CI [–0.002, 0.077]) nor its indirect effect through goal setting (γ = 0.010, 95% CI [–0.005, 0.030]) on PCSB is significant. Although feedback-based goal setting accounts for 22.2% of the relationship between negative feedback and PCSB, the results do not support H1b because the indirect effect is insignificant.

Interaction effects

Table 2 presents the findings of the model that incorporates interaction effects. Initiative climate moderates the relationship between feedback-based goal setting and PCSB (γ = 0.360, 95% CI [0.011, 0.708]). The slope figure (Web Appendix C, Fig. WC.1) illustrates a stronger relationship when the level of initiative climate is high (γ = 0.449, 95% CI [0.277, 0.619]) than when it is low (γ = 0.159, 95% CI [–0.010, 0.330]). The region of significance for the initiative climate is 2.27 (i.e., 0.27 standard deviations below the mean value of 2.54). The slope of the feedback-based goal setting–PCSB relationship is significant for initiative climate values above 2.27 but not below 2.27. These findings support H2.

Table 2 Results of the interaction effects model (Study 1)
Full size table

Additional analyses and robustness check

Additional analyses confirm the robustness of the proposed model, supporting its configuration and the hypothesized relationships (for more details, see Web Appendix D).

Re-estimating the model with OLS

We employed ordinary least squares (OLS) to re-estimate the model, introducing additional control variables such as work unit and hotel dummies. The results are consistent with the random-effects model, confirming the internal validity of our findings.

Feedback as a composite variable

We tested the effect of feedback on feedback-based goal setting using a composite variable that combines positive and negative feedback. The combined variable was not significantly related to goal setting, underscoring the importance of distinguishing between positive and negative feedback to gain a more comprehensive understanding of their unique impact on the process.

Moderating role of feedback accuracy

We do not find a significant moderating effect of perceived feedback accuracy on the relationship between positive/negative feedback and goal setting. However, the significant direct effect of feedback accuracy on goal setting underscores its importance in the goal-setting process regardless of feedback valence.

Alternative mediating variable

The results show that feedback-based goal setting is not related to ERB. However, ERB has a significant effect on customer service performance, even after controlling for PCSB. Additionally, the interaction of initiative climate with feedback-based goal setting has no significant effect on ERB. These findings confirm that ERB and PCSB are distinct constructs with unique effects on customer service performance.

Serial mediation

The results indicate a significant serial mediation effect: positive feedback is linked significantly to customer service performance through feedback-based goal setting and PCSB (see Web Appendix C, Table WC.6).

Alternative models

Alternative models in which we tested the moderating role of initiative climate in the relationship between positive/negative feedback and feedback-based goal setting and in the relationship between PCSB/ERB and customer service performance show no moderating effect.

Study 2

Purpose

The purpose of Study 2 is twofold. First, the results of Study 1 support a positive relationship between positive feedback and feedback-based goal setting. However, the effect of negative feedback on feedback-based goal setting was not significant. Therefore, we conducted an experiment to examine the underlying mechanism operating between positive/negative feedback and feedback-based goal setting to understand why negative feedback did not have a significant effect. Additionally, we explored the differences in motivation resulting from positive and negative feedback in the goal-setting process. Second, by manipulating feedback valence in an experimental setting, we aimed to increase the causality of our findings, which can be problematic with survey data, despite the time-lagged design used in Study 1. The research question we sought to answer in Study 2 was, “What are the processes by which positive and negative feedback impact feedback-based goal setting?” To address this, we tested the mediating role of feedback utility and feedback accountability between positive/negative feedback and feedback-based goal-setting (see Web Appendix E, Figure WE.1).

The mediating effect of feedback utility and feedback accountability

By providing information about successful and failed actions, positive and negative feedback motivate goal pursuit in different ways. Positive feedback increases motivation by signaling goal attainment, while negative feedback increases motivation by signaling insufficient goal progress. Similarly, we posit that positive and negative feedback affect goal setting via feedback utility and accountability, respectively. Feedback utility is “the tendency to believe that the feedback is useful in achieving goals” (Linderbaum & Levy, 2010, p. 1376). Based on the goal-setting and feedback literature, we propose that employees utilize positive feedback in their goal setting for the following reasons. Positive feedback serves as motivation and reinforcement for employees, shaping their beliefs about instrumentality (i.e., efforts will lead to other desirable outcomes) and expectancy (i.e., the probability that the outcome will be achieved) (Fishbach et al., 2010; Fishbach & Finkelstein, 2012). By providing feedback on accomplishments and strengths, positive (vs. negative) feedback increases self-efficacy and elicits positive feelings (Bandura, 1991; Fishbach et al., 2010). Thus, employees regard positive evaluations as reinforcing and are likely to use this information (Brett & Atwater, 2001). Therefore, feedback utility mediates the relationship between positive (vs. negative) feedback and feedback-based goal setting.

In contrast, we propose that negative (vs. positive) feedback enhances feedback accountability by eliciting awareness of insufficient goal progress. Feedback accountability is “the tendency to feel a sense of obligation to react to and follow up on feedback” (Linderbaum & Levy, 2010, p. 1376). By highlighting insufficient accomplishment, negative feedback emphasizes the discrepancy between an employee’s current and desired performance (Fishbach et al., 2010). It alerts them to the discrepancy and prompts them to reduce this gap to meet their goals (Fishbach & Finkelstein, 2012). While unpleasant, negative feedback promotes goal pursuit by making employees feel obliged and accountable to act on feedback to close the gap between their current and desired performance. Therefore, feedback accountability mediates the relationship between negative (vs. positive) feedback and feedback-based goal setting. In summary, we argue that positive feedback reinforces continued effort through validation, while negative feedback prompts corrective action through accountability.

H3a

Feedback utility mediates the effect of positive (vs. negative) feedback on feedback-based goal setting.

H3b

Feedback accountability mediates the effect of negative (vs. positive) feedback on feedback-based goal setting.

Method

Design and sample

We conducted an online experiment with a single-factor between-subjects design (negative vs. positive feedback). We recruited 298 participants on Prolific for monetary compensation. Of the participants, 60.1% were female, the average age was 30.9 years (SD = 10.1), and the average work experience in a customer-facing job role was 11.00 years (SD = 15.72). The sectoral distribution of the participants was as follows: retail (57.4%), tourism and hospitality (39.6%), and other (3%).

Procedure

Participants were asked to imagine receiving a performance appraisal from their manager. In the negative (vs. positive) feedback condition, participants were informed that they underperformed (vs. performed well), scoring at the bottom (vs. top) 10th percentile. They were also told they needed to improve (vs. did well on) their customer service quality and responsiveness (see Web Appendix E for the stimuli). After reading their performance appraisal, respondents rated feedback utility and feedback accountability along with feedback-based goal setting. As a manipulation check, they assessed the type of feedback given (“What type of managerial feedback did you receive?” 1 = “negative,” 2 = “positive”). We measured feedback utility and feedback accountability with a three-item, seven-point Likert scale (Linderbaum & Levy, 2010). We used a four-item, seven-point Likert scale to measure feedback-based goal setting (Renn & Fedor, 2001). We report details of the measurement model in Web Appendix E.

Results

Manipulation check

As we expected, participants in the positive (vs. negative) feedback condition perceived their feedback as more positive (Mpositive = 6.33, SD = 0.87 vs. Mnegative = 1.95, SD = 1.95; t(296) = − 34.28, p < .001; Cohen’s d = − 3.97).

Feedback utility and accountability

As we predicted, participants in the positive (vs. negative) feedback condition reported higher levels of feedback utility (Mpositive = 5.43, SD = 1.07 vs. Mnegative = 5.17, SD = 1.16; t(296) = − 2.01, p < .05; Cohen’s d = –0.23). Participants in the negative (vs. positive) feedback condition reported higher levels of feedback accountability (Mnegative = 5.48, SD = 1.06 vs. Mpositive = 5.20, SD = 1.01; t(296) = − 2.33, p < .01; Cohen’s d = –0.27).

Mediation analysis

We conducted a mediation analysis using the PROCESS (Model 4) module with bootstrapping (20,000 samples) (Hayes, 2022). Positive (vs. negative) feedback is related significantly to feedback utility (b = 0.260, SE = 0.129, 95% CI [0.007, 0.514]), and feedback utility is related significantly to feedback-based goal setting (b = 0.529, SE = 0.072, 95% CI [0.391, 0.675]). Although the direct effect of positive (vs. negative) feedback on feedback-based goal setting is not significant (b = 0.025, SE = 0.105, 95% CI [–0.179, 0.230]), its indirect effect through feedback utility is significant (b = 0.137, SE = 0.074, 95% CI [0.003, 0.292]). Feedback utility accounts for 84.8% of the relationship between positive (vs. negative) feedback and feedback-based goal setting. Together, these findings support H3a, which predicts that feedback utility is a full mediator in this relationship.

Negative (vs. positive) feedback is related significantly to feedback accountability (b = 0.280, SE = 0.120, 95% CI [046, 0.512]), and feedback accountability is related significantly to feedback-based goal setting (b = 0.299, SE = 0.079, 95% CI [0.142, 0.448]). Negative (vs. positive) feedback has no significant direct effect on feedback-based goal setting (b = –0.025, SE = 0.105, 95% CI [–0.228, 0.182]), but its indirect effect through feedback accountability is significant (b = 0.084, SE = 0.045, 95% CI [0.010, 0.184]). Feedback accountability mediates 77.2% of the relationship between negative (vs. positive) feedback and feedback-based goal setting.Footnote 3 These findings support H3b, which predicts that feedback accountability fully mediates the relationship between negative (vs. positive) feedback and feedback-based goal setting.

Study 3

Purpose

The purpose of Study 3 is twofold. First, while Study 2 elucidated the processes by which feedback impacts feedback-based goal setting, Study 3 tests the generalizability of this effect for different feedback formats and contexts. Specifically, we show that (a) feedback accountability mediates the effect of negative (vs. positive) feedback on goal setting and (b) feedback utility mediates the effect of positive (vs. negative) feedback on the goal setting of service employees in the insurance services industry. In addition to percentile scores, we include written feedback in our feedback manipulations to mimic performance appraisals. Second, because most service employees are given mixed feedback in practice, we examine the effects of mixed feedback on feedback accountability, feedback utility, and feedback-based goal setting.

Method

Design and sample

We conducted an online experiment with a single-factor between-subjects design (mixed vs. positive vs. negative feedback). We recruited 279 participants on Prolific for monetary compensation. To replicate Study 2 in a different context, we restricted recruitment to participants from the insurance services industry. Of the participants, 36.6% were female, the average age was 36.72 years (SD = 9.66), and the average work experience in a customer-facing job role was 12.65 years (SD = 11.83).

Procedure

Participants were asked to imagine receiving a performance appraisal from their manager. In the mixed (vs. positive vs. negative) feedback condition, participants were informed that their performance was average (vs. outstanding vs. poor), scoring at the 50th (vs. top 90th vs. bottom 10th ) percentile. They were also given written feedback on various aspects of their job, such as customer service quality, responsiveness, and product and service knowledge (see Web Appendix F for the stimuli). After reading their performance appraisal, respondents rated feedback utility and feedback accountability along with feedback-based goal setting. As a manipulation check, they assessed the type of feedback given (“What type of managerial feedback did you receive?” − 5 = “negative,” 5 = “positive”). We assessed feedback utility, feedback accountability, and feedback-based goal setting using the same scales as in Study 2. We report details about the measurement model in Web Appendix F.

Results

Manipulation check

As we expected, participants in the positive (vs. negative vs. mixed) feedback condition perceived their feedback as more positive than their counterparts in the negative feedback condition (Mpositive = 4.11, SD = 1.31 vs. Mnegative = − 3.81, SD = 1.68; t(276) = 34.53, p < .001; Cohen’s d = 5.05) and the mixed feedback condition (Mmixed = 0.65, SD = 1.68; t(276) = 14.95, p < .001; Cohen’s d = 2.21).

Feedback utility and accountability

Replicating the findings from Study 2, participants in the positive feedback condition reported higher levels of feedback utility than those in the negative feedback condition (Mpositive = 5.81, SD = 0.83 vs. Mnegative = 4.52, SD = 1.53; t(276) = 6.98, p < .001; Cohen’s d = 1.02) and the mixed feedback condition (Mmixed = 4.62, SD = 1.31; t(276) = 6.39, p < .001; Cohen’s d = 0.94). Participants in the negative feedback condition reported higher levels of feedback accountability than those in the positive feedback condition (Mnegative = 5.66, SD = 0.95 vs. Mpositive = 5.35, SD = 0.87; t(276) = − 2.38, p = .02; Cohen’s d = –0.34). There were no significant differences between the negative and the mixed feedback conditions (Mmixed = 5.47, SD = 0.85; t(276) = –0.96, p = .35; Cohen’s d = –0.15).

Mediation analysis

We conducted a multi-categorical mediation analysis using the PROCESS (Model 4) module with bootstrapping (20,000 samples) (Hayes, 2022) with feedback utility and accountability as parallel mediators. Positive (vs. negative) feedback has a significant effect on feedback utility (b = 1.289, SE = 0.185, 95% CI [0.925, 1.652]), and feedback utility has a significant effect on feedback-based goal setting (b = 0.452, SE = 0.045, 95% CI [0.362, 0.541]. The direct effect of positive (vs. negative) feedback on feedback-based goal setting is significant (b = –0.299, SE = 0.141, 95% CI [–0.576, –0.021]), as is its indirect effect through feedback utility (b = 0.582, SE = 0.114, 95% CI [0.374, 0.820]). Feedback utility mediates 66.1% of the relationship between positive (vs. negative) feedback and feedback-based goal setting. These findings suggest that feedback utility partially mediates the relationship between positive (vs. negative) feedback and feedback-based goal setting.

Positive (vs. mixed) feedback has a significant effect on feedback utility (b = 1.188, SE = 0.186, 95% CI [0.822, 1.155]), and feedback utility has a significant effect on feedback-based goal setting (b = 0.452, SE = 0.045, 95% CI [0.362, 0.541]. The direct effect of positive (vs. mixed) feedback on feedback-based goal setting is not significant (b = –0.047, SE = 0.136, 95% CI [–0.315, 0.222]). However, its indirect effect through feedback utility is significant (b = 0.537, SE = 0.116, 95% CI [0.329, 0.782]). Feedback utility accounts for 92% of the relationship between positive (vs. mixed) feedback and feedback-based goal setting, indicating its role as a full mediator.

Negative (vs. positive) feedback has a significant effect on feedback accountability (b = 0.311, SE = 0.130, 95% CI [0.054, 0.568]), and feedback accountability has a significant effect on feedback-based goal setting (b = 0.398, SE = 0.064, 95% CI [0.271, 0.524]. The direct effect of negative (vs. positive) feedback on feedback-based goal setting is significant (b = 0.299, SE = 0.141, 95% CI [0.021, 0.576]), as is its indirect effect through feedback accountability (b = 0.124, SE = 0.060, 95% CI [0.017, 0.253]). Feedback accountability accounts for 29.3% of the relationship between negative (vs. positive) feedback and feedback-based goal setting, indicating its role as a partial mediator.

Negative (vs. mixed) feedback is not related significantly to feedback accountability (b = 0.185, SE = 0.130, 95% CI [–0.072, 0.442]), but feedback accountability is related significantly to feedback-based goal setting (b = 0.398, SE = 0.064, 95% CI [0.271, 0.524]. The direct effect of negative (vs. mixed) feedback on feedback-based goal setting is significant (b = 0.252, SE = 0.123, 95% CI [0.010, 0.495]), but its indirect effect through feedback accountability is not significant (b = 0.074, SE = 0.057, 95% CI [–0.027, 0.199]). Thus, feedback accountability does not mediate the relationship between negative (versus mixed) feedback and feedback-based goal setting.

In summary, Study 3 replicates Study 2 in the insurance services industry. Feedback accountability mediates the effect of negative (vs. positive) feedback on feedback-based goal setting. In contrast, feedback utility mediates the effect of positive (vs. negative and vs. mixed) feedback on feedback-based goal setting. Thus, we demonstrate that feedback valence underlies the relationship between feedback and feedback-based goal setting. Moreover, we elucidate the effects of mixed feedback. Rather than flowing through both mechanisms, mixed (vs. positive) feedback decreases goal setting by reducing feedback utility. Previous research has shown that employees benefit from positive feedback because it highlights personal accomplishments and strengths (Bandura, 1991; Brett & Atwater, 2001; Fishbach et al., 2010). Because mixed feedback is less self-enhancing than purely positive feedback, employees may be less inclined to use this information, reducing its usefulness. Similarly, mixed feedback may include positive evaluations, highlighting a slight discrepancy between an employee’s current and desired performance rather than purely negative feedback. Therefore, mixed feedback may be insufficient to make employees feel obligated or responsible for closing this gap, which may explain the lack of significant effects on feedback accountability.

Discussion

Summary of findings

This paper contributes to the literature by investigating the MG3 model, which synthesizes feedback mechanisms, goal-setting principles, and the PCSB literature into a unified framework. Across three studies, we confirm the validity and robustness of the MG3 model.

In Study 1, we test the entire MG3 model, revealing two key findings: (a) positive feedback exerts its influence on PCSB through feedback-based goal setting, though we do not observe this mediation for negative feedback; (b) the positive impact of feedback-based goal setting on PCSB depends on the presence of a high-initiative (but not a low- initiative) environment. Building on these findings, Studies 2 and 3 further examine the complex relationship between feedback and feedback-based goal setting and provide a nuanced understanding of the underlying mechanisms. In both studies, the findings confirm that feedback utility mediates the effect of positive feedback, while feedback conscientiousness mediates the effect of negative feedback on feedback-based goal setting. Through these three studies, our research offers a comprehensive and holistic perspective by combining three research streams and dissecting the effective mechanisms. Our findings contribute to the advancement of theory, provide valuable practical insights, and highlight potential avenues for further research.

Theoretical contributions

This study makes several noteworthy contributions to the existing literature. First, we offer initial empirical evidence demonstrating that feedback-based goal setting mediates the relationship between positive feedback and PCSB. Despite PCSB being inherently a deliberate and purposeful goal-driven process (Parker et al., 2010), this crucial relationship has not been empirically examined before. As Crant (2000) argues, “It is the confluence of individual differences, contextual factors, and perceptual sense-making through mediating and moderating processes that ultimately determines one’s propensity to engage in proactive behavior.” By establishing this connection, we contribute to the feedback, goal-setting, and PCSB literature, shedding light on a critical pathway through which feedback can influence PCSB. Our findings underscore the pivotal role of feedback-based goal setting as an intervening variable, particularly in the case of positive feedback. Second, we delve into the underlying mechanisms by examining how positive and negative feedback distinctly shape feedback-based goal setting. While prior literature has explored the mechanisms of positive and negative feedback (e.g., Ilies et al., 2010), our study is the first to empirically demonstrate distinct mediators for these two types of feedback. By enhancing feedback utility, positive feedback reinforces goal commitment and underscores the benefits of receiving feedback from a prescriptive standpoint. Conversely, negative feedback, by emphasizing feedback accountability, draws attention to deficiencies in goal progress and underscores the importance of considering feedback from a normative perspective. Our findings suggest that positive and negative feedback can facilitate feedback-based goal setting and are not mutually exclusive or conflicting. It is imperative to recognize that both types of feedback contribute to feedback-based goal setting irrespective of the conditioning role of feedback accuracy, albeit through different pathways. In doing so, we demystify the “black box” between positive/negative feedback and feedback-based goal setting, thus making an important contribution to the intersection of the feedback and the goal-setting literature.

Third, while prior research has made strides in understanding the formation of initiative climate through initiative-enhancing human resources management systems and its influence on motivational states, such as role breadth, self-efficacy, and intrinsic motivation (Hong et al., 2016), there has been limited exploration of the moderating role of initiative climate in the PCSB process, with a few exceptions (e.g., Sok et al., 2021). Our research sheds light on this by demonstrating that although feedback-based goal setting alone can serve as a potent motivator for employees to engage in PCSB, an unsupportive climate (i.e., one that does not foster self-initiation, taking charge, or anticipation) can impede the impact of feedback-based goal setting on PCSB. Building on Raub and Liao’s (2012) findings, which highlight the positive moderating role of initiative climate in the relationship between general self-efficacy and proactive service performance, our results extend this limited research domain. Specifically, we provide empirical evidence that once an employee establishes goals to improve customer service based on manager feedback, a high initiative climate becomes crucial to realize the full benefits of feedback-based goal setting on PCSB. In essence, our research underscores the significance of a cross-level interaction effect by confirming that feedback-based goal setting, as an individual-level construct, is limited in influencing another individual-level construct like PCSB unless a supportive work environment, such as a high initiative climate, is present. This nuanced understanding contributes to the literature by emphasizing the interplay between individual and contextual factors in the PCSB process.

Fourth, while the marketing literature comprises extensive research on ERB, service-oriented citizenship behavior, service-directed ERB, and broader areas like prosocial behavior (e.g., Bettencourt & Brown, 1997; Bettencourt et al., 2001; Chan & Wan, 2012; Maxham & Netemeyer, 2003; Netemeyer et al., 2005), our research takes a pioneering step by incorporating both PCSB and ERB within a single model. It is essential to recognize that not all PCSBs are discretionary; some may fall within an employee’s prescribed role. This is further supported by Crant (2000, p. 436), who suggests that “employees can engage in proactive activities as part of their in-role behavior in which they fulfill basic job requirements.” Therefore, it is essential to recognize that PCSB can encompass both extra-role and in-role behaviors. Similarly, ERBs can involve both anticipatory and self-initiated actions and reactive responses to specific situations such as responding to customer requests, complaints, or service failures (Raub & Liao, 2012). A substantial shared variance between these two constructs could undermine our model’s and hypotheses’ validity when ERB is introduced. Nevertheless, our hypotheses remain unaltered, reaffirming the conceptual distinction between PCSB and ERB, thereby enhancing the robustness of our model and the credibility of our results. This nuanced approach contributes to the literature by elucidating the boundaries and relationships between these distinct but related forms of workplace behavior.

Though beyond the scope of this research, a potential explanation for the null relationship between negative feedback and feedback-based goal setting could be attributed to differing attitudes toward negative feedback. These may include self-defensiveness, disagreement with the evaluation, self-protection, or discomfort with criticism, leading to lower self-esteem, ego, and social image (Kim & Kim, 2020; Linderbaum & Levy, 2010). Nonetheless, our study responds to calls and enriches the literature by advocating for proactive behavior models that incorporate moderators and mediators, thus offering deeper insights into underlying processes (Parker et al., 2010).

We conducted an additional test to discern the distinct impacts of positive and negative feedback on feedback-based goal setting. We combined service employees’ responses to both feedback forms, irrespective of their valence, resulting in a composite variable. We aimed to scrutinize whether this combined variable notably influenced feedback-based goal setting. Our findings reveal that the composite variable did not significantly affect feedback-based goal setting, underscoring the need to distinguish between positive and negative feedback to understand their unique effects on feedback-based goal setting comprehensively.

Managerial implications

Drawing on the results of an extensive multi-respondent, multilevel, and time wave survey alongside two controlled experiments, we offer practical and actionable strategies for managers seeking to leverage feedback to encourage more feedback-based goal setting and PCSB among employees. To this end, we propose three specific suggestions. First, in line with our initial research inquiry, managers must recognize that facilitating feedback-based goal setting necessitates more than just giving feedback to employees regardless of its valence. Depending on whether the feedback is positive or negative, managers must cultivate a sense of feedback utility in the case of positive feedback and feedback accountability for negative feedback. These two mechanisms are necessary for the mere provision of positive or negative feedback to succeed in achieving its intended purpose.

We propose an approach akin to the “carrot or stick” principle, rooted in different ways of framing feedback to evoke either feedback utility or accountability, contingent on the feedback’s valence. Specifically, designing a feedback system framed around success or gain could be effective in engendering feedback utility, particularly among individuals characterized by self-enhancement motivations and lower self-criticism tendencies. Managers could implement a “carrot” approach, or a reward-oriented feedback system, wherein feedback is presented as, for instance, “Meeting your goal will help with reaching a bonus and enhance opportunities for promotion.” Such framing can trigger feedback utility, thus enhancing feedback-based goal setting and, eventually, PCSB.

Conversely, a feedback system centered on failure or loss may prove effective in instilling feedback accountability, particularly among individuals driven by self-improvement motivations and those who exhibit higher levels of self-criticism. Here, managers can adopt a “stick” feedback system, exemplified by framing feedback as follows: “Failure to meet the goal will reduce the likelihood of achieving your bonus and diminish opportunities for promotion.” Such framing techniques may evoke feedback accountability, thereby enhancing feedback-based goal setting. In essence, these nuanced approaches to framing feedback can facilitate feedback utility in the case of positive feedback and feedback accountability in the case of negative feedback, contributing to an increase in feedback-based goal setting and, eventually, PCSB.

The second actionable recommendation for managers pertains to the role of initiative climate in sustaining the positive influence of feedback-based goal setting on PCSB. Our findings demonstrate that feedback-based goal setting does not affect PCSB in a low-initiative climate. The impact is only noticeable in the case of a high-initiative climate. This finding suggests that managers should take measures to prevent a low initiative climate. Managers can focus on two areas that may help prevent the decline of initiative climate: human resources management practices and leadership. Firms can establish a strong initiative climate by implementing initiative-enhancing human resources management practices in selection, training, performance evaluation, and rewards that support self-initiating behaviors and risk-taking. Proactive behavior involves challenging the status quo, which can carry risks and lead to unintended errors. As a result, a climate that is complementary to initiative might be a psychological safety climate that is more accepting of deviations from norms and instills a greater sense of tolerance. Leadership styles that encourage autonomy (e.g., empowerment, delegation) or transformation can also be vital in fostering an initiative climate. Therefore, while feedback-based goal setting motivates employees to engage in PCSB, an initiative climate can help sustain PCSB once employees have set feedback-based goals.

Third, managers should embed feedback-based goal setting routinely in personnel development and appraisal systems to promote PCSB. Managers should ensure that feedback-based goal setting occurs frequently to adjust and refine goals based on multiple instances of feedback received throughout the year. Employees can then exercise more PCSB by routinely monitoring and updating their goals in line with the feedback they receive. This dynamic practice can motivate employees to work toward goals, as goals change with employee performance and external factors that are beyond their control.

Finally, although managers may provide feedback that is predominantly either positive or negative, in many situations, managers are likely to give a mix of positive and negative feedback. The results from Study 3 support our proposed mediation mechanism: feedback utility and accountability are robust mechanisms for positive and negative feedback, respectively.

Limitations and directions for further research

This research has limitations that provide opportunities for further research. Prior studies have shown that expertise may be critical when deciding whether to provide positive or negative feedback (Finkelstein & Fishbach, 2012). Novices seek positive feedback, while experts seek negative feedback (Finkelstein & Fishbach, , 2012). For example, students in a beginner French class, compared with those in an advanced class, were more interested in feedback on what they did well (positive feedback). In contrast, students in an advanced French class, compared with those in a beginner class, were more interested in feedback on areas where they could improve (negative feedback) (Fishbach & Finkelstein, 2012). Future studies could build on this by examining how to balance positive and negative feedback for employees with different levels of expertise (e.g., newcomers vs. experienced employees).

While our study focuses on feedback within a hierarchical organizational structure, it is essential to recognize that the dynamics of feedback can vary significantly in flat organizations like startups. In a flat organizational setup, which usually features fewer hierarchical layers and encourages employee autonomy, feedback mechanisms differ. Incorporating 360-degree feedback from various stakeholders, including coworkers, supervisors, subordinates, customers, and suppliers, can provide a well-rounded evaluation. This inclusive approach fosters a more comprehensive and equitable feedback process, ensuring a holistic assessment of employee performance. Furthermore, future studies could draw on research on temporal distance and feedback valence (Fishbach et al., 2006; Lieberman & Trope, 1998), which indicates that positive feedback is more effective for distal goals, whereas negative feedback is more effective for proximal goals. Researchers can determine how to mix feedback valence depending on whether an employee sets a near (3–6 months), intermediate (6–12 months), or distant goal (over 12 months).

We also encourage researchers to explore potential moderators that influence employees’ use of feedback for goal setting. Some of these may include, but are not limited to, goal orientation (i.e., learning vs. performance orientation), employment status (part-time vs. full-time), and feedback self-efficacy (i.e., the extent to which people believe they have the competency to use feedback to make improvements/changes). Another area that warrants investigation is the mode of feedback communication. While some of the more common modes include verbal and written delivery, technology has opened the door to artificial intelligence (AI)-generated feedback. The application of AI appears particularly relevant given recent findings, which show that when an offer does not meet expectations, people tend to prefer that offer if it comes from AI rather than a human source. Conversely, if an offer exceeds expectations, the preference shifts toward a human source over AI (Garvey et al., 2023).

In this research, we investigate the effect of manager feedback on employees’ goal-setting. We examine the relationship by measuring manager feedback at Time 1 and goal setting at Time 2. However, the temporal sequence approach may not eliminate concerns about the reverse causality (or simultaneity) of feedback giving and goal setting. For example, setting unrealistic goals may result in negative feedback, while setting accurate and achievable goals may lead to positive feedback. Therefore, it is necessary to examine more clearly whether positive or negative feedback affects goal setting or whether goal setting affects the valence of the feedback received. To mitigate these concerns, researchers should consider using a repeated measures design to observe changes in manager feedback and employee goal setting over time, similar to panel data. Such an approach could help establish a more robust causal relationship and reduce endogeneity concerns.

In Study 1, we tested our model using data from service employees, who rated positive and negative feedback from managers. Future studies should validate our model using objective feedback data. In addition, some scale factor loadings were below 0.70, possibly due to the study’s specific context (hotels in Turkey). Researchers should evaluate the validity and reliability of these scales across various countries and industries.

We manipulated positive and negative feedback to reflect the upper and lower 10th percentiles of performance, respectively. Although this manipulation was adequate for our research, it may be considered extreme and not frequently occur in typical workplace settings. Therefore, future work should investigate the thresholds of positive and negative feedback at which feedback utility and accountability cease to play a mediating role. Also, in experimental research, a control condition typically excludes the treatment to serve as a baseline. However, in our study, which focused on feedback’s impact on employee behavior, omitting feedback was impractical because we measured outcomes like feedback utility, accountability, and the influence on goal setting. These variables inherently require the presence of feedback to be assessed. Therefore, employing a control condition without feedback would not have been viable for our research, as these constructs emerge through receiving feedback. Thus, our design necessitates that all participants receive some form of feedback. Rather than a control condition, future studies might explore how neutral feedback, in addition to mixed feedback, as in Study 3, stacks up against positive, negative, and mixed feedback.

Considering that positive (vs. mixed) feedback positively influences feedback utility whereas negative (vs. mixed) feedback does not significantly affect feedback accountability, researchers might utilize a mental accounting perspective to explore whether negative feedback carries more weight than positive feedback in a mixed feedback situation (Loewenstein & Prelec, 1993). Additionally, future studies could provide insights into effectively delivering positive and negative feedback. For instance, researchers could investigate whether it is more effective to combine positive and negative feedback when the former outweighs the latter or whether it is preferable to separate the two types of feedback, presenting them sequentially, particularly when negative feedback predominates.

Finally, further research is needed on whether managers should collaborate with employees to set goals following feedback. Insights from research on customer participation can provide valuable perspectives on this matter (Dong & Sivakumar, 2017). Self-serving bias suggests that in joint participation, participants tend to claim more credit when the outcome is positive but assume less responsibility when the outcome is negative (Bendapudi & Leone, 2003). Thus, when managers invite employees to participate in collaborative goal setting based on feedback, different implications may arise depending on the nature of the outcome (positive or negative). As such, managers should be mindful of the consequences potentially arising from joint goal-setting efforts.

Conclusion

Across three studies, this research demonstrates the different paths that positive and negative feedback take to impact PCSB. By testing the MG3 framework (Study 1) and portions of it (Studies 2 and 3), we establish that positive feedback affects feedback-based goal setting through a feedback utility route while negative feedback takes a feedback accountability route. Additionally, the relationship between positive/negative feedback and PCSB is more complex than initially thought. We hope this research lays a foundation for further studies on how managers can leverage feedback with different valence to improve customer service performance.