Trust is an essential lubricant in situations of dependence amid uncertainty (Rousseau et al., 1998). In the SE, trust is particularly crucial because of its unique characteristics: C2C interactions between strangers (Belk, 2014), enabled by digital platforms, involving underutilised assets (Gerwe & Silva, 2020). Unlike traditional dyadic trust (Mayer et al., 1995), SE transactions involve three parties, consumer, provider and platform, with consumers trusting peer-providers rather than brand-affiliated ones (Ramphal, 2024). This digitally-enabled sharing among strangers (Schor, 2016; Sutherland & Jarrahi, 2018) entails risks like asset damage or misrepresentation, making consumer willingness to trust strangers foundational to the SE.

The classic trust preconditions, uncertainty, risk and interdependence (Rousseau et al., 1998), are inherent in SE transactions (Mittendorf et al., 2019). Without traditional regulatory enforcement (Ferrari, 2016; Sundararajan, 2019), SE platforms rely on alternative trust mechanisms: interaction-based trust (from consumer ratings and reviews) and institution-based trust. This raises the question of where trust is most effectively built: at micro (individual), meso (brand) or macro (regulatory) levels.

Interaction-based trust develops through technology-enabled consumer ratings (Abrahao et al., 2017) and reviews (Cheng et al., 2019), with platforms institutionalising decentralised feedback into provider reputations. This emergent, peer-driven mechanism challenges whether traditional regulatory trust can be superseded.

Institution-based trust stems from specific institutional arrangements (Bachmann, 2011), where favourable conditions enable situational success (McKnight & Chervany, 2001). Regulators create this through independent standards-based reputation (Bartlett et al., 2013), while platforms rely on brand assurances like secure transactions and guarantees (Akhmedova et al., 2021; Kong et al., 2020). Consumers expect competent, integrity-based performance within established norms for legitimacy (Mayer et al., 1995; Scott, 2014), making institution-based trust essential alongside interaction-based trust.

Marketing involves institutions and processes that create, deliver and exchange value (American Marketing Association, 2017). Gundlach and Wilkie (2009) highlight marketing’s role in ‘institutions that … facilitate and govern these activities (e.g. governmental agencies, legislators, courts, professional associations, social norms, ethics and individual values)’ (p. 262) and assert ‘that marketing systems … are a part of marketing as are social processes (e.g. regulations and norms)’ (p. 261). Institutions, comprising cultural-cognitive, normative and regulatory dimensions together constrain and enable behaviour (North et al., 2009; Scott, 2014). As an emergent market form, the SE relies on these institutions for legitimacy, shaping trust and participation through norms and values (normative), socially constructed meanings (cultural-cognitive) and rules governing consumption (regulatory) (Scott, 2014).

Therefore, the SE’s institutional landscape and marketing dynamics are co-shaped by consumers, firms, platforms and regulators (Eckhardt et al., 2019). A core feature of this landscape is the consumer’s dual role as a prosumer, simultaneously providing and consuming services. The resulting institutions facilitate value creation (Duncan, 1920, as cited in Rosenbloom, 2013) and coordinate resources across the ecosystem (Edvardsson et al., 2014). By applying institutional theory, this study specifically examines which institutional pillar, normative, cultural-cognitive or regulatory, most effectively influences consumer trust and participation intention in the SE.

Research on the SE typically analyses trust at the micro (individual), meso (platform) and macro (regulatory) levels (Maurer et al., 2020; Trenz et al., 2018). While studies have examined micro-level peer reputations and meso-level platform assurances, macro-level regulatory influences remain underexplored (Mao et al., 2020; Yang et al., 2019). This fragmented approach highlights a need for integrated multi-level theorisation (Cheng, 2016). Trust in this ecosystem is inherently layered – originating in peer interactions (micro), shaped by platforms (meso) and contextualised by regulation (macro) – requiring a unified model to explain its development across interdependent levels (Figure 1).

Institution-based trust in the SE may arise through different reputation mechanisms operating at distinct levels. Platform reputation emerges endogenously through repeated peer-to-peer interactions and is embedded within the digital platform. It is typically reflected in aggregated user ratings and reviews, such as buyer-seller feedback on online marketplaces or passenger ratings of drivers in ride-hailing platforms, which function as socially constructed signals based on prior exchanges (Mauri et al., 2018; Pavlou, 2002). Platform reputation, therefore, reflects normative pressures generated and reinforced through peer communities.

In contrast, independent reputation originates exogenously from third-party institutions outside the platform’s governance structure. Such mechanisms are commonly associated with formal regulation, certification or standard-setting bodies, including quality assurance certifications or licensing requirements, and are grounded in rule-based assessments (Scott, 2014; Zucker, 1986). Independent reputation thus reflects regulatory pressures, whereby trust is conferred through compliance with externally defined checks and balances.

This distinction is central to the present study, which examines how different institutional pillars substitute for trust under uncertainty in digitally mediated markets (Eckhardt et al., 2019). Specifically, it compares the influence of platform-based and independent reputation mechanisms on consumer trust and participation intentions in the SE.

At the micro-level, consumer trust is rooted in the expectation that a service provider will act in the consumer’s best interest, a set of expectations conceptualised as trusting beliefs and intentions, which form the basis of interpersonal trust (McKnight & Chervany, 2001). In the virtual context of the SE, where transactions are initiated before physical service delivery, this translates into interaction-based trust. This form of trust is predicated on the perceived ability, integrity and benevolence of the provider, a framework established in e-commerce (Pavlou & Fygenson, 2006) and extended to SE contexts (Tussyadiah & Park, 2018; Yang et al., 2019). The inherent information asymmetry of interacting ‘at a distance’ presents a central challenge. Platforms mitigate this by institutionalising the aggregated experiences of past users into a platform reputation, enabling consumers to base their trust on this normative, peer-derived signal.

Rather than explicitly assessing the provider’s ability, integrity and benevolence, platforms fabricate a proxy for interaction-based trust by aggregating consumers’ ratings (Abrahao et al., 2017) to infer the provider’s reputation (Vavilis et al., 2014). Consumers rely on ratings, even from those they have not engaged with. As a normative institution, platform reputation reflects shared norms (Scott, 2014) and serves as a consumer decision-making heuristic (Cialdini et al., 1991), with reviews and ratings shaping provider reputations (Mao et al., 2020; Mauri et al., 2018; Yang et al., 2019). However, ratings themselves vary in degree and interact with the platform’s overall reputation. This leads to our first hypothesis:

Brand trust has evolved from traditional mechanisms like insurance and escrows (Zucker, 1986) to corporate brands (Möhlmann & Geissinger, 2018), reflecting consumer confidence in reliability and intent (Delgado-Ballester et al., 2003). In digital contexts, brands signal quality and engender trust through structural assurances (McKnight & Chervany, 2001; Shankar et al., 2002). Similarly, SE platforms build institution-based trust through website reliability, transaction security, fair policies, verification and reduced user friction. These features enhance the platform’s brand appeal, aid user acquisition and retention (Akhmedova et al., 2021) and reinforce trust in its ability and integrity (Mayer et al., 1995). The platform brand thus functions as a cultural-cognitive institution through which consumers develop taken-for-granted expectations (Scott, 2014), as seen when brand names become verbs (e.g. ‘Ubering’). Hence, we hypothesise:

Institution-based trust also operates at the macro-level through standards, regulations and certifications (Lawrence, 1999; Shao et al., 2020). Independent accreditation institutionalises credibility by substituting for reputation and alleviating information asymmetry (Bartlett et al., 2013; Pavlou, 2002; Martin-Fuentes et al., 2018). Official star ratings enhance reputation and trust (Kang et al., 2016; Sutherland et al., 2021), reflecting how normative institutions can evolve into regulatory ones over time (Scott, 2014). Since regulations are externally enforced and carry coercive power, they may exert a stronger influence on trust. This leads to our third hypothesis:

Given the potential for regulatory ratings to function as mere ‘hygiene factors’ rather than key discriminants, we ask whether trust in formal regulators is as influential as trust in platform reputation systems (Eckhardt et al., 2019). Institutional arrangements often combine multiple elements (Scott, 2014), and contrasting independent regulatory ratings with peer-derived platform ratings offers a novel way to examine how macro- and micro-level trust signals interact. Thus, we hypothesise:

Key drivers behind consumers’ changing perceptions, intentions and actions are evolving cultural, institutional and technological norms (Zhang & Chang, 2020), which have permeated consumers’ lives through the SE. Consumers’ intention to participate in the SE can be defined as their likelihood to request or use a sharing service (Mittendorf et al., 2019). Sellers with better reputations tend to attract more customers (Tadelis, 2016). Consumers’ trust in providers shapes their intentions in e-commerce (Fang et al., 2014), accommodation sharing (Mao et al., 2020; Nisar et al., 2020) and ride-sharing (Mittendorf et al., 2019). Our fifth hypothesis is therefore:

When a provider’s rating was experimentally manipulated, higher values signalled a greater willingness to use the service (Rosenthal et al., 2020). Additionally, higher ratings indicated that providers would remain in the market, thus serving as a proxy of consumers’ support of participation in the SE service (Leoni, 2020). Our sixth and seventh hypotheses, therefore, accommodate both participation intention and the role of trust as a mediator:

Correspondingly, consumers value standard-setting as it leads to more bookings and greater market share (Ballina et al., 2020):

Consumer intention to participate in the SE is predicated on trust in the platform itself. This platform trust is multifaceted, encompassing dimensions such as safety measures, guarantees and perceived service quality, which collectively constitute the platform’s brand (Ter Huurne et al., 2017). Empirical research substantiates that trust in the platform is a key driver of participation intentions (Lee et al., 2018; Mittendorf, 2018; Mittendorf et al., 2019). Consequently, a platform’s ability to foster adoption hinges on reinforcing this brand-based trust (Akhmedova et al., 2020). This establishes the foundational logic for our final hypothesis:

Figure 2 represents the conceptual model of the hypotheses.

The structural model tested five direct hypotheses, of which four were supported: H1, H2, H5, and H8 (see Table 5 for complete results). H3 was not supported, indicating that IR does not function as hypothesised.

Analyses of PR confirmed its strong normative influence. Compared to the 3-star reference, a 1-star PR rating significantly reduced TSP, supporting H1a. Conversely, a 5-star PR rating produced a stronger positive effect on TSP (H1b). This asymmetric pattern underscores that consumers’ interaction-based trust is highly sensitive to peer ratings, with perfect scores providing a substantial boost and low scores imposing a severe penalty.

Supporting H2, B exerted a significant positive influence on TP. This finding confirms that the brand acts as a critical cultural-cognitive institution, providing the structural assurance necessary for institution-based trust at the meso-level.

The results for IR were mixed and revealing. While a 1-star IR rating significantly reduced TSP (H3a), a 5-star rating had no statistically significant effect (H3b). This pattern indicates that a poor regulatory signal can damage trust, but a positive one does not comparably enhance it. This asymmetry suggests that relative to powerful peer-based signals (PR), macro-level regulatory assurances (IR) have a limited and non-linear role in building consumer trust, with consumers appearing to rely more heavily on collective peer experiences than on formal ratings.

Finally, both TSP and TP were confirmed as significant drivers of participation intention (IP), supporting H5 and H8, respectively. The standardised path coefficient for TSP (β = 0.564) was substantially larger than that for TP (β = 0.201), indicating that trust in the individual provider is a more decisive factor in the participation decision than trust in the platform itself.

The mediation analysis, detailed in Table 6, examined the indirect pathways to participation intention (IP). The results indicate significant mediation for most hypothesised paths (H6 and H9), with one key exception: the effect of a high (5-star) IR rating in H7.

Platform reputation (PR) as a mediated pathway: The analysis for H6 confirmed a significant indirect effect of PR on IP through TSP. For the 1-star PR rating (versus the 3-star reference), both the direct and indirect paths to IP were negative, indicating complementary mediation. This suggests that a poor peer rating not only directly deters participation but also does so by eroding trust in the provider. Conversely, the 5-star PR rating showed significant positive direct and indirect effects, also demonstrating complementary mediation, whereby an excellent peer rating boosts both trust and intention directly.

The limited, mediated role of independent reputation (IR): The results for H7 were more nuanced. The direct effects of both low and high IR ratings on IP were non-significant (p = 0.922 and p = 0.727, respectively). However, for the 1-star IR rating, a significant full mediation was observed (β = −0.368, p = 0.001). This indicates that a negative regulatory signal only influences participation intention by first undermining TSP, with no direct effect. The non-significant mediation for the 5-star IR rating reinforces that positive regulatory signals have no reliable effect – neither direct nor indirect – on participation.

Brand trust as a mediating mechanism: Supporting H9, B exhibited a significant positive indirect effect on IP through TP, demonstrating full mediation. This confirms that the platform’s brand does not influence participation intention directly; its effect is entirely channelled through the trust it fosters in the platform itself.

The analysis of the interaction between PR and IR (H4) reveals a nuanced pattern in how these signals jointly shape TSP, as detailed in Table 7 and visualised in Figure 4.

The effect of PR was contingent on its level. As hypothesised in H1a and H1b, 1-star PR ratings (TCs 1, 4 and 7) consistently reduced TSP, while 5-star PR ratings (TCs 3, 6 and 9) consistently enhanced it, regardless of the accompanying IR level. Critically, medium (3-star) PR ratings (TCs 2, 5 and 8) yielded non-significant coefficients, suggesting that average peer ratings are insufficient to build trust and may be functionally equivalent to a lack of information.

The role of IR was more complex and context-dependent. A 1-star IR rating (TCs 1, 2 and 3) exerted a uniformly negative influence on TSP, supporting H3a. However, contrary to H3b, higher IR ratings (3- and 5-stars) showed positive coefficients across conditions, though their impact was not uniformly strong enough to achieve significance in the main effects model. This indicates that while a poor regulatory signal is damaging, a positive one is not a consistently powerful trust signal on its own.

The interaction is clearly illustrated in Figure 4. The slope representing the relationship between PR and TSP is steepest when paired with a 5-star IR rating, indicating that regulatory endorsement amplifies the positive effect of high peer ratings. Conversely, a provider with a 1-star PR rating suffers a severe trust deficit that even a perfect 5-star IR rating (TC 7) cannot fully offset. In contrast, the penalty for a low IR rating is less absolute; it lowers trust but does not negate the powerful positive effect of a 5-star PR score. This pattern confirms the primacy of peer-based (normative) signals, while revealing that macro-level (regulatory) signals can act as a moderating amplifier, particularly for high-quality providers.

Figure 5 summarises results, greying out unsupported hypotheses (H3 and H4) and highlighting partial mediation (H6).