The servicescape—the physical setting where service encounters take place—includes ambient conditions, spatial layout and functionality, signs, symbols, décor, and other environmental cues ([5]). When evaluating a service experience, customers read features like lighting, music, scent, color, cleanliness, employee appearance, and facility layout ([5]). Well-designed environments foster pleasure, comfort, and approach behaviors, while poorly designed ones tend to produce avoidance ([15]). [28] showed that when customers perceived greater value in a small business’s servicescape, they also reported higher levels of loyalty, a finding that positioned the servicescape as a potentially important relational resource for small businesses.

Consumer responses to service environments are shaped not only by physical characteristics but also by customers’ psychological needs and motivations. [34] defined need for affiliation as an individual’s desire for acceptance, belonging, and positive interpersonal relationships. This perspective aligns with broader motivational theories that emphasize the importance of social connection and belonging ([4]; [32]). Individuals higher in nAff therefore place greater value on environments that facilitate social interaction, interpersonal comfort, and relationship development.

Servicescapes can help meet these needs because physical surroundings communicate social meanings alongside functional ones. Features like seating arrangements, ambient conditions, décor, and employee presentation foster perceptions of warmth, hospitality, and community ([36]). [44] similarly argue that servicescapes can function as social environments that promote belonging and emotional attachment. These qualities matter especially in small businesses, which often distinguish themselves through personalized service, community engagement, and close customer relationships.

Consistent with this reasoning, [28] found that customers higher in nAff rated the perceived value of small business servicescapes more favorably. These findings suggest that servicescapes are not interpreted uniformly; rather, customers’ underlying affiliation motives shape the value they assign to environmental and social cues. This relationship provides the motivational starting point for the present framework, which examines how favorable servicescape evaluations contribute to trust and loyalty.

Trust is central to relationship marketing because it reduces uncertainty and gives customers confidence that an organization will behave reliably in future exchanges ([37]). [33] proposed that trust develops through evaluations of ability, benevolence, and integrity. Ability refers to perceptions of competence, benevolence reflects concern for the other party’s interests, and integrity represents consistency with accepted principles and commitments. These judgments shape an individual’s willingness to rely on another party within an exchange relationship.

Empirical research consistently links customer trust with favorable relational outcomes. Customers who trust organizations report greater satisfaction, commitment, willingness to recommend, and intent to maintain ongoing relationships ([10]; [19]; [47]). More recent work reinforces this pattern across service contexts, showing that trust reduces perceived risk and strengthens confidence in future interactions ([25]; [61]). For small businesses, where customers often interact directly with owners and employees and have fewer external sources of information about organizational reputation, trust may matter even more.

Servicescapes may contribute to trust because customers rely on observable environmental features to infer organizational qualities they cannot directly assess. Signaling Theory holds that individuals turn to observable signals when information asymmetry prevents them from judging underlying quality or reliability firsthand ([48]). In service settings, cleanliness, maintenance, layout, décor, ambient conditions, and employee presentation convey competence, professionalism, and operating standards ([5]). Contemporary servicescape research echoes this view, suggesting that physical and social environments shape customers’ assessments of organizational quality and credibility ([29]; [58]).

These environmental signals can also be read through [33] dimensions of trustworthiness. A well-maintained, professionally designed environment signals ability, while customer-focused layouts and welcoming surroundings convey benevolence. Consistency in maintenance, cleanliness, and presentation likewise builds perceptions of integrity. Such cues stand out especially for small businesses, which often depend on direct customer experience rather than established corporate reputations to earn legitimacy and credibility.

A significant amount of empirical work supports a link between service environments and trust: Research across hospitality, retail, healthcare, financial services, and digital commerce shows that favorable environmental evaluations tie to greater confidence, lower perceived risk, and stronger relationship quality ([21]; [31]; [53]). These findings suggest that customers read servicescapes as information about the business itself, which provides a basis for examining trust as an intervening mechanism between perceived servicescape value and loyalty.

Figure 1 presents the proposed conceptual model:

Figure 1. Conceptual model.

Individuals high in nAff are driven to seek social connection, belonging, and positive interpersonal relationships ([4]; [35]). As noted above, because servicescapes convey social and relational information through layout, décor, ambient conditions, and employee presentation, affiliation-oriented customers place greater value on environments that foster interpersonal comfort and connection ([5]; [44]). This matters especially in small businesses, where personalized service and community engagement often form core parts of the customer experience. Consistent with [28] finding that customers higher in nAff report more favorable servicescape evaluations, the following hypothesis remains in the model for validation:

H1: Higher need for affiliation is positively associated with perceived value ofservicescapes.

Servicescape evaluations shape customers’ cognitive, emotional, and behavioral responses to service environments ([5]; [36]). Favorable environmental evaluations tie to stronger behavioral intentions and loyalty across service settings ([22]; [57]). Within small businesses, where customer experience plays a key role in differentiation and relationship development, favorable servicescape perceptions similarly encourage ongoing patronage. [28] provide direct evidence of a positive link between perceived servicescape value and loyalty in small businesses. Because the present study examines whether trust explains part rather than all of this relationship, the direct path is also retained in the model:

H2: Higher perceived value ofservicescapesis positively associated with customer loyalty.

As explained above, customers often rely on observable service-environment cues to judge organizational characteristics that are hard to assess directly ([5]; [48]). A favorable servicescape can signal competence and professionalism, while customer-focused environmental features convey concern for customer well-being. These signals map onto the ability, benevolence, and integrity dimensions that underlie trust development in the Integrative Model of Organizational Trust ([33]). Existing empirical work further shows that service environments are linked to relationship-relevant outcomes, including service evaluations, attachment, trust, and behavioral intentions ([21]; [31]; [53]). Accordingly:

H3: Higher perceived value ofservicescapesis positively associated with customer trust.

Trust reduces uncertainty about future exchanges and increases customers’ willingness to stay in relationships with organizations they regard as dependable and reliable ([33]; [37]). Empirical research has consistently linked trust to customer loyalty and other long-term relational outcomes ([10]; [19]; [47]). This relationship is especially important in small businesses, where ongoing interactions with owners and employees reinforce perceptions of competence, honesty, and dependability. Therefore:

H4: Higher levels of customer trust are positively associated with higher levels of customer loyalty.

These four proposed relationships suggest that trust explains part of the process by which favorable servicescape evaluations develop into loyalty. Signaling Theory holds that customers use servicescape cues to infer organizational characteristics under conditions of uncertainty ([48]), while the Integrative Model of Organizational Trust explains how evaluations of ability, benevolence, and integrity build trust ([33]). Commitment-Trust Theory, in turn, identifies trust as a central mechanism supporting enduring exchange relationships ([37]). Thus, favorable evaluations of a small business’s servicescape strengthen trust, which in turn is associated with loyalty.

This mediation framework extends [28] by moving beyond the direct servicescape-loyalty association to examine a relational mechanism underlying that relationship. Because the model also retains the direct path from servicescape value to loyalty, the analysis can determine whether trust accounts for part or all of the relationship. Accordingly:

H5: Customer trust mediates the relationship between the perceived value ofservicescapesand customer loyalty.

Participants were recruited from customers of selected small businesses in a mid-sized metropolitan area of the Northeast region of the United States. For purposes of the survey, a small business was defined generally as “a privately owned, for-profit company that is independently operated, is not dominant in its field, and typically employs fewer workers and generates lower annual revenue than large corporations” ([1]). At participating businesses, customers were invited at checkout to participate in the survey and were informed that they could enter a drawing for a gift card as an incentive for participation. Eligibility required only that the individual had just patronized one of the participating small businesses where the survey opportunity was made available. Respondents did not independently select a business from a list. Rather, recruitment at a participating small business established the small-business context for the survey, and the survey items asked respondents to evaluate the small business(es) they frequently patronized. Additionally, business-sector information was not collected as part of the study’s anonymity-preserving data-collection procedure. Therefore, the distribution of respondents across specific retail, service, hospitality, or other business sectors cannot be reported.

A total of 333 individuals began the survey but sixty-eight respondents discontinued participation before completing more than half of the relatively lengthy questionnaire and were excluded from the analysis. These cases contained substantial blocks of missing data across the study measures rather than isolated item-level omissions and therefore lacked sufficient information for inclusion in the CFA and SEM. The remaining 265 respondents provided complete data for the variables included in the analyses, resulting in a final analytic sample representing 79.6% of those who began the survey. Because many of the respondents who discontinued the survey did so before providing complete demographic and construct information, a meaningful comparison between included and excluded respondents was not possible. Complete-case analysis was therefore used because the excluded cases contained insufficient information for estimation of the four-construct measurement and structural models. Additionally, because the total number of customers who were invited to complete the survey at the participating businesses was not recorded, a conventional response rate based on the total number of individuals approached cannot be calculated. The typical respondent was between 42 and 57 years old and Caucasian, and 65.5% of respondents were female.

The present study constitutes a secondary analysis of the dataset originally reported by [28]. The original survey included measures of need for affiliation, perceived value of servicescapes, customer loyalty, and customer trust. Although customer-trust items were collected during the original data collection, they were not incorporated into the analyses reported by [28], whose analytical scope focused on the relationships among need for affiliation, perceived servicescape value, and customer loyalty. The prior publication stated that trust data were not collected; more precisely, trust measures were collected but were not analyzed or reported in that study. The present study therefore uses these previously collected trust measures in a secondary analysis addressing a theoretically distinct research question.

The analytical contribution of the present study extends beyond adding customer trust as an additional construct to the earlier model. Whereas [28] examined whether need for affiliation was associated with perceived servicescape value and whether servicescape value was associated with customer loyalty, the present study examines the relational mechanism through which favorable servicescape evaluations may contribute to loyalty. Specifically, the analysis integrates Signaling Theory ([48]), the Integrative Model of Organizational Trust ([33]), and Commitment-Trust Theory ([37]) to examine whether favorable servicescape evaluations are associated with greater customer trust and whether trust mediates the relationship between perceived servicescape value and customer loyalty.

Sample adequacy was further evaluated using semPower 2.2.0 ([38]). Because the analysis used parameter estimates from the final model, it was treated as a sensitivity analysis rather than as a prospective power calculation. For the structural relationships, the analysis focused on the smallest standardized path coefficient in the model, the relationship between perceived servicescape value and customer loyalty (β = .218), while accounting for customer trust as an additional predictor of loyalty (β = .415) and the association between perceived servicescape value and customer trust (.529). The analysis incorporated four customer-loyalty indicators, 11 perceived-servicescape-value indicators, and seven customer-trust indicators, with average standardized factor loadings of .775, .782, and .821, respectively. At α = .05 and a target power of .80, the estimated minimum sample size required to detect the weakest structural relationship was 224.

A separate latent-variable mediation power analysis evaluated the indirect relationship between perceived servicescape value and customer loyalty through customer trust. The analysis specified the standardized relationships between perceived servicescape value and customer trust (β = .529), customer trust and customer loyalty (β = .415), and the direct relationship between perceived servicescape value and customer loyalty (β = .218), together with the corresponding measurement-model characteristics. At α = .05 and a target power of .80, the estimated minimum sample size for detecting the indirect effect was 66. The final analytic sample of 265 therefore exceeded the estimated sample-size requirements for both the weakest structural relationship and the mediation effect.

All constructs were measured using previously established scales adapted where necessary to fit the small-business context. The adaptations were intended to preserve the conceptual meaning of the original constructs while ensuring that respondents could evaluate the small businesses they patronized and, where applicable, the physical and interpersonal characteristics of those businesses. All items were measured using a five-point Likert-type response format with endpoints ranging from completely agree to completely disagree. For the CFA and SEM analyses, the Likert-type indicators were entered as numeric observed variables and treated as continuous. The complete wording of the retained indicators is provided in Tables A1-A4. Measurement-model refinement and the criteria used to eliminate indicators are described in Section 5.3.

5.2.1. Need for Affiliation

Need for affiliation (nAff) was initially measured using five items from [49]), a measure that continues to be used in contemporary research examining individual manifest needs ([45]; [60]). The scale assesses an individual’s tendency to seek interpersonal involvement, maintain positive relationships with others, and attend to the interpersonal aspects of social situations. Following a confirmatory factor analysis (CFA), three indicators satisfied the established loading criterion and were retained.

The retained indicators represent complementary manifestations of the underlying need for affiliation rather than separate dimensions of the construct. One item captures preference for working or interacting with others rather than independently, a second reflects attentiveness to the feelings of other people, and a reverse-coded item reflects the tendency to avoid openly expressing interpersonal disagreement. Collectively, these indicators operationalize nAff as an orientation toward interpersonal involvement, sensitivity to others, and relational harmony. Because nAff was modeled as a single latent construct, these item characteristics are described as aspects of construct coverage rather than as independent subdimensions. The wording of the three retained indicators is presented in Table A1.

5.2.2. Perceived Value of Servicescapes

Perceived value of servicescapes was initially measured using 18 items adapted using the perceived-value framework developed by [50]. Their PERVAL framework conceptualizes consumer value as multidimensional, incorporating emotional value, social value, quality/performance value, and price/value-for-money considerations. Because Sweeney and Soutar’s original measure was designed primarily to evaluate consumer products, the present study contextualized the perceived-value framework to the service environment rather than reproducing product-specific wording. Respondents were therefore asked to evaluate the value associated with observable characteristics of the small-business environments they patronized.

The adaptation focused on the experiential and functional value generated by servicescape characteristics. Visual elements such as artwork, wall décor, color, and furnishings captured customers’ aesthetic and quality evaluations. Ambient characteristics such as temperature, aroma, and lighting captured the comfort, pleasure, warmth, and welcoming qualities associated with the service environment. Spatial-layout items addressed functional aspects of value, including crowding and ease of movement. The original 18-item adaptation also contained three employee-related indicators intended to incorporate the social component of the servicescape. Price/value-for-money items were not included because the focal construct concerned the perceived value of the service environment itself rather than the economic value of the overall purchase transaction.

After CFA refinement, 11 of the 18 indicators were retained. Conceptually, the retained indicators primarily represent the emotional and quality/performance components of perceived value identified by [50], expressed through three servicescape-related areas: visual/aesthetic value, ambient/experiential value, and spatial/functional value. The three employee-related social-servicescape indicators did not meet the loading criterion and were removed during measurement-model refinement. Accordingly, the final latent construct should be interpreted primarily as customers’ perceived value of the physical servicescape, rather than as a comprehensive measure of both the physical and social servicescape. Table A2 presents the retained indicators and identifies the principal servicescape component and perceived-value emphasis represented by each item.

5.2.3. Customer Trust

Customer trust was initially assessed using eight items adapted from [13]. The original trust framework distinguishes between reliability and intentions. Reliability concerns confidence that the organization will consistently meet expectations and provide dependable performance, whereas intentions concern customers’ beliefs that the organization will act honestly and constructively when problems or unexpected circumstances arise.

The scale was adapted by replacing the original brand-specific referent with the small businesses patronized by respondents while preserving the substantive meaning of the original items. Following a CFA refinement, seven indicators were retained. Four retained indicators represent the reliability component of trust by assessing whether the business meets expectations, inspires confidence, avoids disappointing customers, and stands behind customer satisfaction. Three additional indicators represent positive customer-oriented intentions by assessing honesty and sincerity in responding to concerns, willingness to solve customer problems, and efforts to satisfy customers. Thus, the retained items capture both dependable performance and positive intentions toward customers. The seven retained customer-trust indicators are presented in Table A3.

5.2.4. Customer Loyalty

Customer loyalty was initially measured using five items adapted from [6]. Although Bobâlcă et al. conceptualize loyalty as a broader process involving cognitive, affective, conative, and action-related stages, the items incorporated into the present survey were drawn from the affective-loyalty component of their scale. The items were adapted from a brand context to the small businesses patronized by respondents.

Once CFA refinement was conducted, four indicators were retained. These indicators assess pleasure in choosing the focal small businesses over alternatives, greater liking of those businesses relative to competing firms, stronger attachment to the businesses, and greater interest in those businesses relative to alternatives. Consequently, customer loyalty in the present study is operationalized primarily as affective or attitudinal loyalty, reflecting favorable preference, attachment, and relational interest rather than observed repeat-purchase behavior. This distinction is important because the study assesses customers’ psychological loyalty toward the businesses they patronize rather than objective behavioral measures such as purchase frequency or transaction history. Table A4 provides the wording of the four retained indicators.

5.2.5. Common Method Bias

Because all constructs were measured through self-reported data collected within a single survey, procedural and statistical safeguards were used to address the possibility of common method bias (CMB). Consistent with [42], respondents were assured of anonymity and item sequencing was varied within the survey. These procedures were intended to reduce evaluation apprehension and respondents’ ability or motivation to provide systematically consistent answers across constructs.

CMB was also assessed statistically using Harman’s single-factor test. All measurement items were entered into an unrotated exploratory factor analysis. The first unrotated factor accounted for 35.40% of the total variance, below the commonly applied 50% threshold. This result suggests that a single common factor did not account for the majority of covariance among the measures. Nevertheless, because the study employed a cross-sectional self-report design, the possibility of residual common method variance cannot be completely excluded.

To provide a more stringent assessment of common method variance, a CFA-based single-factor model was also estimated in which all 25 retained indicators were constrained to load on a single latent factor. The single-factor model demonstrated poor fit to the data (χ² = 2106.58, df = 275, CFI = .598, TLI = .561, IFI = .600, GFI = .466, AGFI = .368, RMSEA = .159). In contrast, as discussed below, the proposed four-factor measurement model demonstrated substantially better fit. The difference between the models was statistically significant, Δχ²(6) = 1624.341, p < .001. These results indicate that a single latent factor does not adequately account for the covariance among the study measures and provide additional evidence against a dominant common method factor.

In addition, a common latent method-factor sensitivity analysis was also conducted to assess whether shared measurement method materially affected the substantive relationships in the model. A common method factor was specified to load on all 25 retained indicators while remaining orthogonal to the substantive latent constructs. The method-factor model demonstrated acceptable fit (χ² = 414.06, df = 246, CFI = .963, TLI = .955, IFI = .964, GFI = .890, AGFI = .854, RMSEA = .051). Although several indicators exhibited some method-related variance, the substantive structural coefficients remained highly stable after inclusion of the method factor. Specifically, the nAff → perceived servicescape value coefficient changed from .326 to .317, perceived servicescape value → customer trust from .529 to .524, perceived servicescape value → customer loyalty from .218 to .204, and customer trust → customer loyalty from .415 to .429. The largest absolute change in any structural coefficient was .014. These results indicate that the substantive relationships were robust to the inclusion of a common latent method factor and were not materially attributable to shared method variance.

Confirmatory factor analysis (CFA) and structural equation modeling (SEM) were conducted in AMOS 28.0 using maximum likelihood (ML) estimation. Prior to model estimation, the retained indicators were examined for distributional characteristics. Univariate skewness values ranged from −.568 to −.003, and kurtosis values ranged from −.997 to −.397, indicating no evidence of problematic univariate nonnormality. AMOS’s assessment of multivariate normality produced a multivariate kurtosis value of 19.191 with a critical ratio of 4.251, indicating some departure from multivariate normality. Following [51] recommendations, indicators with standardized factor loadings below .71 were removed. Consistent with prior scale-refinement procedures ([7]; [23]), 11 items were eliminated, leaving 25 indicators with standardized loadings ranging from .716 to .865, as shown in Table 1. The retained indicators continued to represent the theoretical domains of nAff, perceived servicescape value, customer trust, and customer loyalty.

Table 1. Measurement model summary.

Eleven items with standardized loadings below .71 were removed following the initial CFA: two need for affiliation items, seven perceived value of servicescapes items, one customer trust item, and one customer loyalty item.

The revised measurement model demonstrated acceptable fit (χ² = 482.24, df = 269, CFI = .953, TLI = .948, IFI = .954, GFI = .873, AGFI = .846, RMSEA = .055). Composite reliability (CR) exceeded .70 and average variance extracted (AVE) exceeded .50 for all constructs, supporting reliability and convergent validity ([18]), as shown in Table 2. The retained indicators also loaded significantly on their intended constructs, providing further evidence of convergent validity ([2]).

Table 2.Composite reliability and average variance extracted.

Composite reliability values above .70 indicate satisfactory construct reliability; AVE values above .50 indicate adequate convergent validity.

Discriminant validity was assessed using the [18] criterion. For each construct, the square root of the average variance extracted (AVE) exceeded its correlations with all other constructs in the measurement model, indicating that the constructs share more variance with their own indicators than with each other. This is shown in Table 3; the pattern provides evidence of adequate discriminant validity.

Table 3. Discriminant validity assessment.

The bold diagonal values are the square roots of AVE. In every case, the diagonal value is larger than the correlations involving that construct, so the Fornell-Larcker criterion supports discriminant validity.

Structural equation modeling (SEM) using AMOS 28.0 was employed to test the hypothesized relationships among need for affiliation, perceived value of servicescapes, customer trust, and customer loyalty. The structural model demonstrated acceptable fit to the data (χ² = 507.61, df = 271, χ²/df = 1.873, CFI = .948, TLI = .943, IFI = .948, GFI = .866, AGFI = .840, RMSEA = .058). The χ²/df ratio was below the recommended value of 3.0, providing additional evidence of acceptable model fit ([9]).

The structural path estimates supported H1 through H4. Need for affiliation was positively associated with perceived value of servicescapes (β = .326, p < .001), supporting H1. Perceived value of servicescapes was positively associated with customer loyalty (β = .218, p = .002), supporting H2, and with customer trust (β = .529, p < .001), supporting H3. Customer trust was also positively associated with customer loyalty (β = .415, p< .001), supporting H4. Thus, customers with stronger affiliation needs tended to report greater perceived servicescape value, while favorable servicescape evaluations were associated with both greater trust and greater loyalty.

H5 proposed that customer trust mediates the relationship between perceived value of servicescapes and customer loyalty. Given the observed departure from multivariate normality, the indirect effect was evaluated using nonparametric bootstrapping in AMOS with 5000 resamples and bias-corrected 95% confidence intervals. The standardized indirect effect was positive and statistically significant (β = .219), with a bias-corrected 95% confidence interval that did not include zero (95% CI [.143, .308]). H5 was therefore supported. Because the direct relationship between perceived servicescape value and customer loyalty remained significant after customer trust was included in the model (β = .218, p = .002), the results indicate complementary mediation ([62]). Favorable servicescape evaluations were therefore associated with customer loyalty both directly and indirectly through increased customer trust.

Figure 2 presents the structural model with standardized path estimates; Table 4 presents the summary of the hypothesis testing results:

Figure 2. Structural model and standardized path estimates.

Table 4.Summary of hypothesis testing results.

β represents standardized path estimates (for H5, β represents the standardized indirect effect).