Relationships among actors within the sharing economy: Meta-analytics review

The sharing economy has emerged as an influential research area in which a platform mediates customers ’ temporary access to service provider resources. To provide a generalizable picture of the platform ’ s customer and service provider relationship formation process, we integrate effect sizes from 192 studies, including 214 independent samples (N = 88,154). The findings indicate there are motivators and inhibitors for individuals to join a platform as a customer or service provider and that these influence attitudinal and behavioral responses toward the platform through a two-level relationship quality pathway. Moderator analysis reveals that the impact of customer motivators and inhibitors on customer response to service providers and platforms depends on country-level moderators and cultural context. These results provide insight into relationship formation among actors in the sharing economy. The study also makes recommendations for platform managers, especially in hospitality and tourism, to more effectively manage their relationships with their users.


Introduction
The sharing economy business model became popularized with two Silicon Valley start-ups, Airbnb and Uber (Eckhardt et al., 2019).The economic situation in most countries, changes in consumer behavior, and rapid development in technology have accelerated the success of the sharing economy (Benoit et al., 2017) and encouraged the purveyors of various products and services to adopt this business model (Kumar et al., 2018).Given the impressive growth of the sharing economy, it is not surprising that this business model has been heralded as a global transformation that has a significant influential impact on the global economy (Eckhardt et al., 2019).
With regards to its importance, many researchers have studied the different aspects of relationship formation in the sharing economy (Hamari et al., 2016).Initial research on the sharing economy focused on factors that motivate individuals as customers (e.g., guests of an Airbnb) and service providers (e.g., host of an Airbnb) to join the sharing economy and on risks related to participation in the sharing economy (Benoit et al., 2017).A review of sharing economy literature indicates that authors have proposed a wide range of factors that facilitate or impede customers and service providers regarding joining a sharing platform (Kumar et al., 2018).However, there is no agreement about the number and nature of these motivators and inhibitors (Cheng, 2016;Hamari et al., 2016).
Another research area requiring further attention in the sharing economy is relationship quality (Breidbach and Brodie, 2017).Some studies have examined relationship quality at the platform level (Arteaga-Sánchez et al., 2018;Lee Zach et al., 2018), comparing it to customer-firm relationships in the traditional business model, looking at the relationship developed by both customer and service provider with the platform.While the sharing economy is a triadic business model, there is a dual interaction among customers, service providers, and platforms (Apte and Davis, 2019;Mittendorf et al., 2019).Therefore, relationship quality in the sharing economy requires study at both the customer-service provider (Mao et al., 2020;Zhang et al., 2018) and platform levels (Yang et al., 2019).
As platforms mediate exchanges between customer and service provider, and their existence depends on customer and service provider loyalty, loyalty is considered a critical research area in the sharing economy (Akhmedova et al., 2020;Arteaga-Sánchez et al., 2018).However, the literature presents a diverse view of customer and service provider loyalty and their intention to remain with a platform.For instance, some researchers studied loyalty at the customer and service provider level (Yang et al., 2017), whereas others investigated loyalty at the platform level (Kumar et al., 2018).
Despite a considerable body of research on the sharing economy in the past decade, there remain debates and disagreements around components of the actor-relationships process (Eckhardt et al., 2019;Perren and Kozinets, 2018).No comprehensive model has been developed to include all actors and their relationship formation process.This study aims to review empirical research on the sharing economy to develop an integrated and comprehensive model of the antecedent, mediators, moderators, and actors' relationship development in the sharing economy.As such, service ecosystem model used for this study includes micro, meso, and macro levels to investigate actor relationship formation in the sharing economy (Akaka et al., 2015;Breidbach and Brodie, 2017).In this model, actors' actions and interactions (i.e., customer and service provider) at the individual level are mediated by an actor at the meso level (i.e., platform), and the macro-level context moderates relationships among actors at micro and meso levels.This model will provide insight into (i) customer and service provider motivators and inhibitors to use the sharing economy, (ii) customer and service provider relationship quality formation at the micro level, (iii) the outcome of the relationship quality among customers, service providers, and platforms at the meso level, and (iv) the role of country-level moderators in these relationships at the macro level.This model will provide theoretical and empirical insight into a unique aspect of relationship formation and development in the sharing economy business model and define areas requiring further research.The following section outlines the model in greater detail.

Conceptual framework
In contrast to traditional business, the sharing economy is a triadic business model that includes three actors, i.e., customer, service provider, and platform (Benoit et al., 2017;Kumar et al., 2018).In this business model, the service provider delivers underutilized resources for customers' temporary access to these resources for a fee (Eckhardt et al., 2019).In this process, a technology intermediary, in the form of a platform, mediates the exchange between two other actors (customer and service provider) for a fee (Kumar et al., 2018).In addition, there are different levels of interaction between actors in this business model.Customer and service provider interactions are mediated through the platform, and institutional logic governs and guides interactions among all these actors.
As there are multiple and multilevel interactions among actors in this triadic business model, we developed the Service Ecosystem model to combine and synthesize diverse research in the sharing economy (Akaka et al., 2015;Breidbach and Brodie, 2017).As shown in Fig. 1, this model is a multilevel network that includes micro, meso, and macro levels (Alexander et al., 2018) in which the service exchange between customer and service provider at the micro level is mediated by the platform at the meso level (Breidbach and Brodie, 2017).In addition, contextual factors at the macro level moderate these interactions at the micro and macro levels (Storbacka et al., 2016).
At the micro level, the sharing economy is considered a network of strangers (i.e., customers and service providers), bringing benefits (motivators) and risks (inhibitors) for its users.Motivators suggest that customers and service providers expect benefits from a sharing economy (Benoit et al., 2017), while inhibitors suggest a perceived risk related to using the sharing economy (Lee, 2020;Lee Zach et al., 2018;So et al., 2018).Over time these antecedents constitute customer and service provider relationship quality (i.e., satisfaction and trust) (Hamari et al., 2016).M. Barari et al.While some researchers have considered relationship quality as a general construct, similar to a customer-firm relationship (Arteaga-Sánchez et al., 2018;Lee Zach et al., 2018), triadic interaction among actors allows us to study relationship quality at two levels, micro and meso (Apte and Davis, 2019).As indicated in Fig. 1, customer and service provider relationship quality at the micro level is the predictor of relationship quality with the platform at the meso level (Mittendorf, 2017;Yang et al., 2018).Moreover, at the meso level, customer and service provider relationship quality components influence their positive responses toward the platform (Yang et al., 2019), such as customer loyalty (Breidbach and Brodie, 2017) and service provider retention (Hua et al., 2020).
In the service ecosystem model, moderators at the macro level are contextual variables that can explain inconsistencies in relationships between actors in the micro and meso levels.It is important that there are adequate effect sizes for moderators to be included (Palmatier et al., 2006).Thus, the moderators that have been included in the service ecosystem model are country-level moderators such as Gross Domestic Product (GDP) per capita, Human Development Index (HDI), and culture.Variables such as sample characteristics (student vs. non-student) and document status (published vs. unpublished) are considered as control variables in the model to ensure the variabilities in effect sizes are not because of these variables (Blut and Wang, 2019;Gremler et al., 2019).The following section considers each level of the service ecosystem model in greater detail.

Customer motivators and inhibitors
Traditionally, hedonic and utilitarian values are considered the customer's expected benefits from interaction with a firm (Babin et al., 1994;Gremler et al., 2019;Hennig-Thurau et al., 2002).However, prior research indicates that customer benefits of the sharing economy are not limited to this value (Kumar et al., 2018).Although there is no overarching agreement among researchers, hedonic, utilitarian, social, and environmental factors are considered customer motivators to use the sharing economy (Benoit et al., 2017;Hamari et al., 2016).
The utilitarian or economic value indicates product and service ability to satisfy fundamental customer needs in the exchange (Babin et al., 1994;Holbrook and Hirschman, 1982;Voss et al., 2003).As the sharing economy allows a customer to satisfy basic needs at a lower price (Kim and Jin, 2020), it is considered the main customer driver for using a sharing economy service (Benoit et al., 2017;Trenz et al., 2018).Hedonic value is considered as customer pleasure and fun during the purchase and consumption process (Babin et al., 1994;Holbrook and Hirschman, 1982;Voss et al., 2003).In addition to utilitarian value, hedonic value is essential for customers to participate in the sharing economy (Hamari, 2017).Moreover, the sharing economy also provides an opportunity for the customer to meet new people and interact with them (Eckhardt et al., 2019).Interactions between customers and service providers are at the heart of many sharing economy platforms such as Uber, Airbnb, and TaskRabbit (Benoit et al., 2017).Finally, the environmental value represents customer sustainable resource consumption through access instead of ownership (Hamari et al., 2016).Increasing customer awareness of environmental issues encourages customers to use a sharing economy model rather than traditional business models to protect the environment (Cohen and Kietzmann, 2014).
While a considerable body of research has focused only on the perceived benefits of the sharing economy (Hamari et al., 2016;Kumar et al., 2018), this business model also has its own perceived risks.In this case this indicates a customer's subjective belief that there is some probability of suffering a loss in pursuit of the desired outcome (Mittendorf, 2017).A service provider as a stranger is an independent actor in the sharing economy who is not a trained employee.This can lead to higher service variability and inconsistency in this business model (Lee, 2020).Therefore, risk is an integral part of the sharing economy business model and is a customer inhibitor to joining and using platform services (Lutz et al., 2018).

Customer trust of and satisfaction with service provider
Relationship quality in the sharing economy is derived from customer evaluation of the benefits compared to the risks (i.e., satisfaction and trust) (Benoit et al., 2017) and is studied at both micro and meso levels.At the micro level, customers evaluate satisfaction and trust by comparing what a service provider promised and what they received (Oliver, 1980).While prior research mainly has focused on customer expected benefits from the sharing economy (i.e., utilitarian, hedonic, social, and environmental) as customer satisfaction and trust drivers (Arteaga-Sánchez et al., 2018;Möhlmann, 2015), perceived risk plays an essential role in the customer relationship quality formation (Mao et al., 2020).Customers in the sharing economy not only receive services from a service provider as a stranger but also interact with a potentially different service provider in each interaction (Huurne et al., 2017).Therefore, received benefits could enhance the quality of customer relationships with service providers, while the perceived risk is considered an inhibitor for customer relationship formation (Eckhardt et al., 2019).

Service provider motivators and inhibitors
There is no agreement in prior research about the benefits of the sharing economy for service providers (Hua et al., 2020).Some research has pointed to the same benefits for customers and service providers (Hamari et al., 2016).However, a service provider is an independent economic actor in the sharing economy who likes to exchange resources on a platform for a fee (Benoit et al., 2017;Kumar et al., 2018).In this regard, prior research indicates that economic value, work flexibility, and social value are the service provider's primary motivators to join and stay with a platform (Benoit et al., 2017;Hua et al., 2020;Trenz et al., 2018;Yeager et al., 2020).
Economic value is the service providers' initial motivation for joining the sharing economy (Trenz et al., 2018).For example, a service provider may seek to get extra income from sharing their underutilized resources, such as a room, car, skills, or tools (Benoit et al., 2017).From this view, the sharing economy turns individuals into micro-entrepreneurs to make money from their resources (Shiu-Li and Shu-Yu, 2020).A second benefit for a service provider is work flexibility, defined as arrangements that help service providers adjust their volume, timing, and location of work (De Menezes and Kelliher, 2017).The sharing economy as an open business model allows service providers to plan different aspects of their work (Nawaz et al., 2019) and enjoy the flexibility and autonomy of working as a freelancer (Benoit et al., 2017;Shiu-Li and Shu-Yu, 2020).Social interaction and networking with other people are also benefits for service providers participating in the sharing economy (Nawaz et al., 2019).For example, the sharing economy provides an opportunity for service providers to meet new people (customers) and enjoy the social benefits of interactions with their customers (e.g., Airbnb guests or their Uber passengers) (Benoit et al., 2017;Shiu-Li and Shu-Yu, 2020).
Sharing underutilized resources in the sharing economy also brings potential risks that act as inhibitors for service providers (Teubner and Flath, 2019).For instance, with Uber or Airbnb, service providers need to furnish services to a stranger in their own house or car, which may cause damage or create losses (Chen Jengchung et al., 2020).Furthermore, service providers need to share their personal information and information about their resources on public platforms, which creates risk for them (Teubner and Flath, 2019).Thus, perceived risk plays a vital role in the service provider experience of working in a sharing economy (Teubner and Flath, 2019).

Service provider relationships quality
Service provider relationship quality in our model is reflected by service providers' satisfaction with the sharing economy, which indicates their evaluation of benefits and risks related to working in the sharing economy (Gleim et al., 2019).Prior research in the sharing economy has mainly focused on the benefits of the sharing economy for a service provider (i.e., economic, flexibility, and social value) that determine service provider satisfaction with the sharing economy (Bucher et al., 2016;Hua et al., 2020).However, the sharing economy brings benefits and risks for service providers who share their resources with customers as strangers (Benoit et al., 2017).Therefore, the risks are considered service provider inhibitors that negatively impact their satisfaction toward the sharing economy.

Customer and service provider relationship
Based on partner effects theory, individuals are verbally and nonverbally influenced by other people's characteristics, behavior, or perceptions (Van Dolen et al., 2002).Research in the traditional business model indicates interactions between customers and front-line employees play an essential role in customer perception and further behavioral intention toward an employee (Kumar and Pansari, 2016).This is because employees who are happy with their work tend to deliver excellent service to customers and share their positive emotions (Brown and Lam, 2008;Hogreve et al., 2017).While this is an undeveloped research area in the sharing economy, service provider satisfaction with the sharing economy is manifested in provider behavior with customers, influencing customer satisfaction with service providers and the platform (Ruan, 2020).It is worth mentioning, based on partner effects theory, there are two-way relationships between customer and service provider satisfaction.However, there is not sufficient effect size for customer satisfaction on service provider satisfaction relationship.Therefore, we did not include this direction in our model.The following section explores the meso level.

Meso level 2.2.1. Customer relationships with platform
While the relationship quality in traditional business models is studied at the firm level (Hennig-Thurau and Klee, 1997), the triadic nature of relationships in the sharing economy allows researchers to study relationship quality at both micro and meso levels (Ta et al., 2018;Yang et al., 2019).While a two-level customer relationship quality and the relationship between the levels has not been granted much consideration in sharing economy research (Mao et al., 2020), we consider it a unique feature of this business model.From this view, customers interact with different service providers, and their satisfaction and trust over time spills over to the platform and determines the quality of customers' relationships with the platform (Eckhardt et al., 2019;Moon et al., 2019).
Customers' relationship quality with the platform determines their loyalty toward the platform (Möhlmann, 2015).Although it is possible to study customer loyalty at both service provider and platform levels, customers receive services from different service providers for each service encounter in a platform such as Airbnb, but they cannot specifically request the same service provider again in a platform such as Uber (Eckhardt et al., 2019).In this regard, loyalty has been studied at the platform level in the sharing economy, in which customers' relationship quality with the platform influences their tendency to use the platform again in the near future (Kong et al., 2020;Ye et al., 2019).

Service provider relationships with platform
Service providers' retention indicates their tendency to stay with a platform and continue working in it (Hogreve et al., 2017;Shiu-Li and Shu-Yu, 2020).While the relationship between service provider and platform has not received considerable attention from researchers, some studies have considered the service provider to be an employee of the platform (Hua et al., 2020).In contrast to traditional business models, a service provider could work through several platforms at the same time.For instance, a driver could work on Uber and/or Didi and/or Ola simultaneously.As service providers are independent actors, their satisfaction with other actors (i.e., customers) spills over to the platform and determines their tendency to stay with it (Lin et al., 2020).The final level to consider is macro.

Macro level 2.3.1. Gross national product
As the gross national product (GDP) reflects people's purchasing power, it may have an impact on customer decision-making (Berry et al., 2010).Customers in countries with a lower level of GDP have less disposable income and seem to prefer the sharing economy as a cost-effective way to access products and services rather than buying them (Blut and Wang, 2019).Therefore, it is expected that the relationship between customer motivators for the sharing economy and customer responses is stronger in countries with a lower level of GDP than those with a higher GDP level (Parente et al., 2018).From a risk perspective, customers from countries with a higher level of GDP are more sensitive to risk in the sharing economy, and, therefore, GDP has a higher impact on customer response (Blut and Wang, 2019).

Human development index
The Human Development Index (HDI) measures country achievements in different areas such as long and healthy life, education, and standard of living (Nations, 2018).From a consumption perspective, countries with a higher level of HDI have more knowledge of and experience with new technologies, such as the sharing economy platforms, compared to countries with lower levels of HDI (Blut and Wang, 2019).This knowledge and experience helps customers in countries with a higher level of HDI enjoy sharing economy benefits, which has an impact on customer attitudinal and behavioral responses to the sharing economy services.In contrast, a lack of knowledge of and experience with the sharing economy in countries with a lower level of HDI increases customers' perceived risk of the sharing economy service usage.

Cultural context
Hofstede et al. ( 2005) have developed a popular approach for studying differences between countries from a cultural perspective.Based on this approach, cultural difference is reflected in four important dimensions: power distance, individualism, masculinity, and uncertainty avoidance.Culture plays an important role in the entire customer shopping process and is considered an important moderator in meta-analysis studies (Orsingher et al., 2010;Van Vaerenbergh et al., 2018).In this regard, previous research indicates that people from different cultures have differing views on the sharing economy's expected values and potential risks, and these differences influence customer responses to these motivators and inhibitors (Albinsson et al., 2019).Following previous meta-analyses (Blut et al., 2016;Pick and Eisend, 2016), cultural dimensions (i.e., power distance, individualism, masculinity, and uncertainty avoidance) have been selected as cultural moderators.

Data collection and coding
Similar to prior meta-analyses (Gui et al., 2020;Park and Min, 2020), we followed a comprehensive approach to identify all potential publications in the field of hospitality and business.We used keywords such as "sharing economy," "access-based consumption," "collaborative consumption," "peer to peer consumption," "peer to peer lending," "peer to peer economy," "access economy," "collaborative economy," and "peer economy" in popular online databases including ABI/INFORM Global, Business Source Complete, ProQuest Digital Dissertations, Scopus, SSRN, Emerald, Springer, ISI Web of Science, and Taylor & Francis.We selected 2010 to the present as the time frame as the appearance of the sharing economy in hospitality and business publications predominantly occurs from 2010 onwards (Botsman and Rogers, 2010).Moreover, to ensure all related articles were included in our data set, we manually checked the titles and abstracts of articles published in top journals in the hospitality industry (e.g., International Journal of Hospitality Management, International Journal of Contemporary Hospitality Management) and in business (e.g., Academy of Management Journal, Journal of Applied Psychology).Finally, 875 publications were considered for further analysis.
In the next step, we defined several inclusion criteria for empirical research studies in our meta-analysis.First, the sharing economy is a common research area in different disciplines, and its definition may differ.Thus, we excluded empirical research in which the sharing economy is related to not-for-profit platforms (e.g., Couchsurfing) and research featuring buying and selling platforms (e.g., eBay.com) as part of the sharing economy, which led to the exclusion of 236 studies.In addition, we included only empirical research that reported correlation matrices or other statistical information (e.g., standardized regression coefficients, t-values) that we could use to calculate a correlation coefficient for a desired relationship (Park and Min, 2020).In this step, we excluded 447 publications.Also, researchers may conduct several studies in a single research endeavor to analyze the same relationship in their conceptual models and report multiple effect sizes for this relationship.If these effect sizes are from independent samples, we included them as separate effect sizes; otherwise (Hunter and Schmidt, 2004), a procedure was used to calculate a composite correlation.Therefore, the final data set included 192 studies with 214 independent samples and a total sample size of 88,154 that met all our criteria.
The coding manual was first developed for coding studies to provide the details of the main and moderator variables in our conceptual model (see Table 1) and reduce the discrepancy in the coding process.Two people were involved in the coding process: one of the authors and an independent coder.The first coder was responsible for coding the studies, and an independent coder who is an expert in the sharing economy area and not involved in this research checked the coding quality.In this regard, 20% of studies were randomly selected, and each coder coded them separately.The overall inter-coder agreement was higher than 95%, confirming coding quality.Differences in coding were resolved through discussion.

Meta-analytic procedures and analysis
A random-effect meta-analysis method was conducted to synthesize effect sizes (Hunter and Schmidt, 2004).As most empirical studies in the sharing economy report correlation, correlation coefficients were used to calculate the effect size.For studies that did not report the correlation coefficient, the data available in the study (e.g., standardized regression coefficients or t-values) was used to calculate the correlation coefficient (Peterson and Brown, 2005).In the next step, correlations were corrected for measurement error: each correlation was divided by the square root between variables of interest reliabilities (Hunter and Schmidt, 2004).Then, the reliability-adjusted correlations of each study were weighted with their corresponding sample size.Also, we calculated the 95% confidence intervals to determine the statistical significance of effect size and the 80% credibility intervals to measure the variability of effect size across studies (Park and Min, 2020).In addition, Hedges's Q statistic was used to test effect size homogeneity.Significant Q-statistics indicate variance in effect size distribution and point to the necessity for moderation analysis (Grewal et al., 2018).
To test the study's conceptual model, we employed meta-analysis structural equation modelling (SEM), allowing the researchers to assess different conceptual models to find the superior model in a domain (Grewal et al., 2018).Reliability-adjusted and sample size-weighted correlations from 192 studies with 214 samples were used to create a pooled correlation matrix (Barari et al., 2021).Then, the matrix was used as a SEM input to simultaneously test the relationships between research variables in our conceptual model (Grewal et al., 2018).
A multilevel meta-regression approach was employed to test the role of moderators in our conceptual model (Hox, 2010), because this method accounts for the dependency between effect sizes from the same sample (Blut and Wang, 2019) and provides a more accurate estimation (Gremler et al., 2019).Following Hox (2010) guidelines, effect sizes were considered as the dependent variable while mediators and outcome variables were predictors in level 1.Independent variables for level 2 included moderators and control variables.For moderators, HDI (Nations, 2018), GDP per capita (Fund, 2020), and the four cultural values (Hofstede et al., 2005) were used as continuous values for the model.Control variables including sample type (student sample = 1 versus non-student = 0) and publication status (published research = 1 versus unpublished = 0) were entered in the model as dummy coded variables.

Descriptive statistics
Descriptive analysis of the relationship between research variables is illustrated in Table 2.The results indicate that there are positive and significant correlations between research variables in our model, with the exceptions of customer perceived risk-satisfaction with a service provider (ρ = − 0.34), customer perceived risk-trust of service provider (ρ = − 0.35), and service provider perceived risk-satisfaction with sharing economy (ρ = − 0.31) where there are negative and significant correlations.For most correlations, Hedges's Q statistic results are significant, indicating the heterogeneity between effect sizes.In addition, the wide difference between the lower and upper bounds of the 80% credibility intervals shows the variance in effect size.Hedges's Q statistic and 80% credibility intervals findings emphasize the necessity of moderator analysis to explain these heterogeneities.
For the service provider, the data analyses indicated that service provider motivators such as economic (β = 0.24, p < .001),flexibility (β = 0.13, p < .001),and social value (β = 0.09, p < .001)have significant and positive impacts while perceived risk (β = − 0.24, p < .001)has a significant and negative influence on service provider satisfaction with the sharing economy.In addition, service provider satisfaction with the sharing economy significantly influenced service provider retention in a platform (β = 0.43, p < .001).Finally, results indicated that service provider satisfaction with the sharing economy is a significant predictor of both customer satisfaction with a service provider (β = 0.20, p < .001)and customer satisfaction with a platform (β = 0.08, p < .001).

Results of moderator analysis
Table 3 shows the results of the moderator analysis between motivators and inhibitors with level one and two variables.
HDI.As predicted, HDI positively and significantly moderated the relationship between motivators and customer responses.Compared to countries with lower levels of HDI, the influence of utilitarian (γ = 0.813, p < .001),hedonic (γ = 0.712, p < .001),social (γ = 0.083, p < .10)and environmental (γ = 0.194, p < .05)values on customer responses are higher among customers from countries with higher HDI levels.In contrast, the impact of inhibitors, i.e., perceived risk (γ = − 0.595, p < .01), on customer response is stronger for countries with a lower level of HDI.
Control variables.The control variables analysis indicated no significant patterns for the study characteristics.The student sample does not moderate the relationship between motivators and inhibitors and customer responses.Similarly, publication status did not significantly moderate any relationships.

Discussion
Testing our conceptual model allowed us to make several contributions to the sharing economy literature.Additionally, our findings have implications for service providers and platforms marketing managers.We summarize our main research findings and theoretical and managerial implications in Table 4.

Theoretical implications
Our service ecosystem model includes three actors (i.e., customers, service providers, and platforms) to illustrate the multi-actor nature of the sharing economy business model and the multilevel relationships among actors in the sharing economy ecosystem (Breidbach and Brodie, 2017;Fehrer et al., 2018;Storbacka et al., 2016).Our model confirms that customer and service provider relationships at the micro level in this ecosystem influence customer and service provider relationships with a platform at the meso level.These relationships impact customer and service provider responses to the sharing economy platform (Storbacka et al., 2016).Moreover, contextual moderators at the macro level moderate the relationships among actors at the micro and meso levels.
For the customer, the findings confirm the role of the motivators and inhibitors as customer relationship formation initiators in the sharing economy (Benoit et al., 2017;Hamari et al., 2016;Möhlmann, 2015).For motivators, utilitarian, hedonic, social, and environmental values determine the level of customer relationship quality with a service provider.While the relative importance of these values in previous studies are diverse and contradictory (Arteaga-Sánchez et al., 2018;Hamari et al., 2016;Hwang and Griffiths, 2017;Lee Zach et al., 2018), the results indicate that utilitarian and hedonic values have a higher impact on customer satisfaction with, and trust in, the service provider than do social and environmental values (Eckhardt et al., 2019).Besides that, customer motivators studies indicate that hedonic value has a higher impact on customer responses to a firm than do utilitarian values (Barari et al., 2020;Chitturi et al., 2008), and our findings confirm the dominant role of utilitarian value for the customer in the sharing economy (Eckhardt et al., 2019).Similarly, our results support the negative impact of perceived risk as an essential inhibitor on customer relationship formation with service providers as strangers in the sharing economy (Lee Zach et al., 2018;Lutz et al., 2018;Yang et al., 2019).
While many sharing economy models focused merely on the benefits of the sharing economy for customers (Hamari et al., 2016;Kumar et al., 2018), our analysis highlights that perceived risk has a destructive role in customer satisfaction and trust in a service provider (Mittendorf et al., 2019;Teubner and Flath, 2019).Moreover, its negative impact on customer satisfaction and trust is larger in magnitude than the positive effect of both social and environmental value.
Our model highlights the difference between the relationship formation process in the sharing economy and customer-firm relationships in traditional business models (Aurier and N'Goala, 2010;Palmatier et al., 2006).The key difference is the duality of relationship quality (i.e., satisfaction and trust) between actors in the sharing economy (Lin et al., 2019;Mao et al., 2020;Mittendorf et al., 2019;Yang et al., 2019) in which customer-service provider relationship quality spills over to the platform and results in high-quality relationships between customers and the platform (Mittendorf et al., 2019;Ta et al., 2018).Also, a high-quality relationship between customers and the platform leads to customer loyalty toward a platform (Lee Zach et al., 2018).Finally, some studies considered customer loyalty with both service providers and platforms.However, the nature of the relationship between customer

Table 4
The study key findings and their implications.

Key findings
Research and managerial implications

Sharing economy and platform
In sharing economy as a triadic business model, the relationship between customer and service provider at the micro level spills over to the platform at the meso level.
As customers and service providers are independent actors, in contrast to B2C business models, platforms require managing their relationships with both customers and service providers.Customer Among customer motivators, utilitarian and hedonic values have the highest impact on customer-service provider relationship quality.Also, the negative impact of perceived risk is larger than the positive effect of both social and environmental value in relationship quality.
Customer relationship quality occurs in two levels: micro level as customerservice provider relationship quality and meso level as customer-platform relationship quality.Customer-service provider relationship quality through customer-platform relationship quality influences customer loyalty to the platform.
Platforms require to provide enough benefits for customers, especially in the form of hedonic and utilitarian values, to compensate customer risk of receiving service from service providers to enhance their relationship with service providers.Platforms' relationship formation with customers is more complex than B2C business models, where platforms do not have enough control over customerservice provider relationships.
The nature of customer relationships in the sharing economy limits customer loyalty to the platform.

Service provider
Among sharing economy benefits, economic value has the highest impact on service provider satisfaction.The negative impact of perceived risk is larger in magnitude than the positive effect of both flexibility and social value.
Service provider satisfaction plays an important role in customer satisfaction, more important than social and environmental values.Also, their intention to work in the platform is the main driver of their satisfaction with the sharing economy.
Service providers are main drivers of sharing economy business models.Thus, platforms need to define enough benefits, mainly economic, for them to recompense the risk of working in this business model.Platforms need to pay close attention to service providers' satisfaction because it has a dual impact on the sharing economy.It enhances both service providers and their customer relationship with sharing economy platform.

Moderators
The influence of motivators on mediators and outcome variables is stronger in countries with a higher HDI level, while this is the opposite for perceived risk.
The role of motivators on customer relationships and their outcomes is more effective in countries with lower power distance and uncertainty avoidance.
Platforms should focus more on customer knowledge and experience in countries with lower HDI levels to facilitate customer relationships.
To enhance customer relationship formation, platforms need to consider customer risk of receiving service in countries with higher power distance and uncertainty avoidance levels.
M. Barari et al. and service provider (e.g., the matching system in Uber) limits the loyalty study to the platform level (Eckhardt et al., 2019).In this regard, customer-platform relationship quality determines customer tendency to use a platform again (Lee Zach et al., 2018).
As with customer relationships, there are both motivators and inhibitors for service providers to join and use sharing economy platforms (Benoit et al., 2017;Kumar et al., 2018).While there is no agreement in prior research about the benefits of the sharing economy for service providers, our findings confirm the positive effects of economic, flexibility, and social value on service provider satisfaction with the sharing economy (Benoit et al., 2017;Hua et al., 2020).The results indicate economic benefits have the primary role in service provider satisfaction with the sharing economy, while flexibility and social value are less critical in this relationship (Kumar et al., 2018).Also, findings confirm the negative influence of perceived risk on service provider satisfaction with the sharing economy (Arteaga-Sánchez et al., 2018).Perceived risk is a strong predictor for individual satisfaction with working in a platform where the negative impact on service provider satisfaction is larger than the positive effect of flexibility and social values.Moreover, our results indicate that service providers' satisfaction with the sharing economy determines their retention with a platform (Hamari et al., 2016;Mittendorf, 2017).As the sharing economy is an open business model, service providers' tendency to continue working in a platform is an important issue.Our results confirm the crucial role of service provider satisfaction on providers' intention to stay with a platform (Eckhardt et al., 2019).
Although investigations into the relationship between service providers and customer relationship quality are limited our findings highlight the influence of service provider satisfaction on customer satisfaction with service providers and platforms.This aligns with research on the customer-employee relationship in which employee satisfaction with the job is a significant predictor of positive customer response (Hogreve et al., 2017).In the customer-firm relationship, employee satisfaction is the main predictor of overall customer satisfaction (Brown and Lam, 2008;Jeon and Choi, 2012).However, service providers are independent of the platform and other service providers, and their satisfaction influences customer satisfaction with both the service provider and the platform (Moon et al., 2019).
Moderator analysis provides some insight into the role of context in the relationship between customer motivators and inhibitors and customer responses in the sharing economy.Our findings show that, in contrast to GDP, HDI significantly moderates relationships in the sharing economy.In this regard, the influence of motivators on mediators and outcome variables is higher in countries with a higher level of HDI, while this is the opposite for perceived risk.These results suggest that the role of motivators on customer response is stronger for countries with higher levels of HDI, where people have enough experience and knowledge to utilize sharing economy services with a lower level of risk (Pick and Eisend, 2016).Moreover, a higher level of technical infrastructure and regulation to support sharing economy ecosystems in countries with higher levels of HDI facilitates motivators on customer response and reduces the negative role of risk (Parente et al., 2018).
Our findings also show that power distance and uncertainty avoidance, among cultural components, significantly moderate relationships in our conceptual model.These findings confirm the role of cultural differences in customer response in the sharing economy (Albinsson et al., 2019;Gupta et al., 2019).While these studies mainly focused on the role of culture on customer intention to choose a sharing economy platform, our results show the role of motivators and inhibitors on customer responses in different cultural contexts.Our findings indicate that an increase in the power distance and uncertainty avoidance weaken the relationship between motivators and customer response and strengthen the role of risk in these relationships.

Managerial implications
The sharing economy has become a popular business model among different industries, with many entrepreneurs and start-ups adopting this model (Huang and Kuo, 2020).However, only a small number of these platforms succeed in attracting enough customers and service providers to become a sustainable business (Täuscher and Kietzmann, 2017).Our conceptual model provides several insights for platform marketing managers to better understand and manage their relationships with customers and service providers.First, in contrast to traditional business models such as B2C, the sharing economy requires platforms to pay equal attention to customers and service providers.Therefore, it is important for a marketing manager who works in this type of business model to develop an ecosystem that is appealing for both supply and demand sides.This could be a challenging task for a platform.For instance, in ridesharing platforms, lower prices encourage more customers to use a platform because they boost customer utilitarian benefits.However, lower prices for customers mean lower economic value for drivers, and this may encourage drivers to switch to a competitor's platform.
Further, for marketing managers our findings highlight the need to consider the complexity of relationship formation with customers in the sharing economy.First, using the sharing economy brings benefits and risks for customers, and managers need to ensure their platform provides enough benefits for customers to compensate for risk.Also, managers should develop a system to maximize customer expected value, especially hedonic and utilitarian, to facilitate customer-service provider relationship quality.For instance, Airbnb enables customers to rate their received value from the service provider, and this rating is reflected on the service provider profile.This system forces service providers to enhance their service values and helps customers choose optimal service providers.Moreover, marketing managers need to be aware that risk plays an important and dysfunctional role in customer satisfaction and trust of a service provider as a stranger.Thus, as with motivators, a mechanism is required to help customers minimize their risk and enhance their relationship with service providers.For instance, Uber allows customers to share their trips, including driver details, with family and friends to reduce their risk.In addition, based on a two-level relationship quality, a platform needs to make sure customers form a high-quality relationship with service providers.For instance, in the Airbnb platform, customers can share different aspects of their experiences with others.This facilitates customer bonds with service providers and improves the quality of their relationship with platforms over time.This is important because platforms cannot increase customer loyalty without having a strong relationship with customers.
Service providers play an essential role in the sharing economy, especially in industries where several platforms compete.Our findings advise platforms to focus mainly on economic value, then flexibility and social value, to satisfy service providers' expected values of working on a platform.For instance, in the ridesharing industry, several platforms such as Uber, Didi, and Ola compete to attract more drivers to their ecosystem to increase their network and profitability.Thus, a platform in this industry requires the development of a payment system to optimize service provider income by considering platform profits and customer service prices.Moreover, marketing managers should pay close attention to service providers' satisfaction as it has a dual impact on the sharing economy ecosystem.For service providers, a higher level of satisfaction means an increase in their tendency to stay with a platform in the future.More importantly, service provider satisfaction has a direct impact on customer relationship quality with both service providers and platforms.
In the international context, our findings provide insights to marketing managers about their relationship with customers.Our findings indicate marketing managers should consider HDI and adjust their marketing strategy in different countries.For countries in which there is platforms need to focus more on customer knowledge and experience to enhance the role of motivators and diminish the impact of inhibitors on customer attitudinal and behavioral responses.Also, platforms should consider the role of cultural context when seeking to expand their business to new countries.From this angle, platforms need to pay further attention to their relationships with their customers who have a higher level of power distance and uncertainty avoidance in countries where the role of motivators on customer relationship and its outcomes is less effective.

Limitations and further research
Like other research methods, our meta-analysis suffers from some limitations that open avenues for future research.The sharing economy conceptual model relies on prior empirical research on the sharing economy.Studies on the sharing economy mostly take the customers' view and consider their relationships with service providers and platforms.For a service provider, there are a limited number of studies on service provider relationships with customers and platforms; thus, this research area requires further study.For instance, in the relationship between customer and service provider satisfaction, this research only studied the role of service provider satisfaction on customer satisfaction.As there is a two-way relationship between the two actors' satisfaction, future research could investigate the complex relationship between these two actors' satisfaction.As the sharing economy is a triadic business model, it is possible to define two-level relationship quality for both customers and service providers.As there was limited available empirical research on the service provider relationship with customers and platforms, we could not test two-level relationship quality for a service provider.Thus, future research should empirically cover this important research area in the sharing economy.
Moreover, most of the prior studies in the sharing economy area tested the relationship between customer and service provider and their relationship formation with each other and platform.Although the platforms depend on service provider resources, they are responsible for a different aspect of the marketing mix.Therefore, platform marketing activities have an essential role in the actor relationship formation.In this regard, there are opportunities for future studies.For instance, a platform is responsible for advertising to create brand awareness to attract more customers and service providers to join a platform.However, we do not know how platform-level advertising could impact customer and firm relationships with each other and platform.In addition, there are differences between platforms in the pricing system.While a service provider is responsible for pricing in the Airbnb platform, Uber is responsible for pricing for service delivered to customers.However, the impact of different pricing systems, especially at the platform level, require further attention and investigation.
COVID 19 Pandemic has resulted in a paradigm shift in consumer's preferences (Gordon-Wilson, 2021;Yap et.al, 2021;Rayburn et.al, 2021;Kursan Milakovic, 2021;Nayal et.al, 2021;Paul & Bhukya, 2021).For example, many consumers did not use sharing economy platforms such as Uber, Airbnb etc.There are opportunities to examine whether the determinants of sharing economic platforms remain the same or not in the post-pandemic era.CB-SEM and PLS-SEM methods (Dash & Paul, 2021) can be used to study such phenomenon.Theoretical explorations and cross-country studies also would be useful.
Note: K: number of effect sizes; N: cumulative sample size; r: average correlation; ρ: reliability adjusted and sample size weighted correlation; SDρ = standard deviation of corrected correlation; CI = confidence interval; CrI = credibility interval; Hedges's Q statistic.*p < .01

Table 1
Constructs definitions and aliases.

Table 2
Descriptive statistics for the relationship between variables.

Table 3
Results of moderator analysis.
1 To create consistency in the interpretation, the perceived risk effect size was reversed.* * p < .01. * p < .05. + p < .10 and customer response.