Omnichannel Fulfilment, Supply Chain Transparency, And Customer Loyalty: A Signalling Perspective

3. METHODOLOGY

3.1. Research Design and Data Collection

This study used a cross-sectional quantitative design to study the relationships between selected dimensions of fulfilment, perceived supply chain transparency, perceived greenwashing, and customer loyalty in the sustainable beauty sector. Data were collected using a self-administered questionnaire (Appendix A) from Malaysian consumers who had purchased products from sustainable beauty brands through both the brands’ online and physical/offline channels. To establish a consistent focal entity across the survey, respondents were instructed to select one sustainable beauty brand that operated both online and physical/offline sales channels and base all subsequent responses on their experiences with that same brand. Thus, the focal entity in the present study was the beauty brand rather than a third-party retailer through which the brand’s products might also be sold. Data collection took place over four months (March to June 2025).

The survey was shared online through social media networks such as Facebook, Instagram, and WhatsApp. The data were examined for missing answers, inconsistent answers, and straight lines before conducting the analysis. First, the participants were asked questions about their place of residence, history of purchasing beauty products from online and offline stores, and their age before being provided access to the complete questionnaire. If a participant answered in any category other than the acceptable category, they were automatically excluded. The respondents were split into two groups for multi-group comparisons: Gen Z (18-26 years) and Millennials (27-42 years), which is similar to the previous study on generations (Pandey et al., 2025).

Prior to answering the questionnaire, the participants were provided with information on the purpose of the study, the voluntary nature of the study, the right to withdraw at any point during the study, and that their answers would be handled confidentially and anonymously. Informed consent was obtained prior to participation. No personal identification was gathered, and the data were only used for academic enquiry. The participants were also informed that there were no known risks other than those associated with participation in a consumer survey.

Purposive sampling was applied to ensure a sample with sufficient experience of digital and physical channels to evaluate aspects of fulfilment. The total number of responses obtained was 452; however, 21 responses were incomplete, nine were straight-lined, and 12 failed the screening process, which led to a final sample size of 410. In accordance with the guidelines of (Hair et al., 2022), this sample size exceeded the minimum requirement using the ten-times rule and was adequate for PLS-SEM testing.

3.2. Measurement Instrument

Each variable was measured using multi-item scales adapted from the literature. The response scale used was a 5-point Likert scale ranging from 1= strongly disagree to 5 = strongly agree. A pilot test was conducted with 30 participants prior to the final data collection to test for understanding, wording, and clarity of the questionnaire. Because the unit of customer evaluation was the focal beauty brand, all customer-facing measurement items were interpreted with reference to the selected brand and its online and physical/offline channels.

Instead of a single higher-order construct, omnichannel fulfilment quality was assessed based on four first-order constructs: delivery service quality, channel integration, fulfilment process efficiency, and returns management efficiency. The measurement items for the four omnichannel fulfilment dimensions were adapted by the authors from the conceptual logistics dimensions discussed by (Marchet et al., 2018). The mentioned study examined omnichannel logistics from an organisational perspective; therefore, their work was used as the conceptual basis for identifying relevant delivery, channel integration, fulfilment, and returns-management dimensions rather than as a source of directly validated customer-level measurement items. This study subsequently adapted and operationalised these dimensions as customer-perceived measures appropriate to the sustainable beauty retailing context. Accordingly, the items reported in Appendix B should be understood as context-specific adaptations rather than verbatim items directly validated by (Marchet et al., 2018).

Saini and Singh (2020) stated that customer loyalty is a multi-dimensional construct involving attitudinal and behavioural loyalty. Attitudinal loyalty is psychological attachment to the brand and preference, whereas behavioural loyalty is reflected in the tendency to buy and recommend the brand. Customer loyalty was modelled as a reflective construct because the measurement items were intended to capture manifestations of respondents’ underlying loyalty toward the focal beauty brand. The construct encompassed both attitudinal and behavioural intention indicators, such as the intention to continue purchasing, willingness to recommend the brand, and preference for the brand. Therefore, these indicators were treated as reflective manifestations of an underlying loyalty orientation rather than distinct dimensions that jointly formed the construct. Accordingly, the items were expected to covary as expressions of the same underlying customer loyalty disposition, and the construct was assessed using the reflective measurement model criteria reported in this study. Given this perspective, the present study adopted the first-order reflective model of customer loyalty, wherein each indicator was considered a complementary form of the same underlying construct as the customer loyalty model. The measurement model demonstrated satisfactory reliability and convergent validity, supporting this operationalisation.

Perceived supply chain transparency was modified from (Morgan et al., 2023) supplier transparency scale. The items were modified to reflect the perceived transparency of customer supply chains from the customers’ perspective. Perceived supply chain transparency was defined as the perception that customers have about the transparency, accessibility, clarity, and visibility of supply chain-related information and processes. This construct does not measure actual transparency in supply chains but rather the customers’ perceptions of such transparency. The items gauge whether customers feel they have sufficient and easy-to-access information about sourcing, production, delivery, and operational processes.

Items measuring perceived greenwashing were adapted from (Tu et al., 2024) to reflect perceptions of deceptive, extensive, or unverifiable environmental claims. Adaptations and adjustments were made to incorporate sustainability claims in the cosmetics industry. The items were selected to capture perceived exaggeration, misleadingness, insufficient substantiation, or misrepresentation of environmental performance, rather than general distrust of the brand.

Some wording changes were made to the items to be consumer-oriented for this study, as several instruments were originally developed for organisational/firm-level respondents. The changes were limited to contextual and grammatical changes, for example, replacing the organisational self-assessment statements with statements based on customer perception but not changing the conceptual definition of the constructs. The adaptations were designed to allow respondents to evaluate their perceptions of the observed fulfilment experiences of the retailer/brand rather than internal organisational processes. The adapted items were checked to ensure that their meaning, in terms of the construct, was maintained. Content validity was assessed by experts in assessing the adapted items and by pilot testing to ensure the clarity and contextual relevance of the items and that they were conceptually consistent with the original measurement scale. Appendix B provides a summary of these tools.

3.3. Data Analysis Procedure

The model was estimated using SmartPLS 4.0.9.9, which is suitable for complex models with mediation, moderation, and Multi-Group Analyses (MGA). A two-step approach was adopted: first, the measurement model was tested, and second, the path model was evaluated. The mediation effects were interpreted in accordance with (Zhao et al., 2010), who argued that a significant indirect effect without a significant direct effect was indirect-only mediation.

A non-parametric bootstrapping procedure with 5,000 resamples was used to evaluate the significance of the structural relationships. The significance of the direct, indirect, and moderating effects was estimated using a bootstrap analysis. To estimate the moderating effect of perceived greenwashing, the Product Indicator approach in SmartPLS was used, which is suitable for modelling interactions between reflective constructs. The dataset was reviewed for missing and inconsistent responses prior to the analysis. During the data cleaning process, incomplete questionnaires were excluded, and the dataset had no missing values; therefore, data imputation was unnecessary.

The MGA was used to examine the structural path coefficients between Gen Z and Millennials to test for cohort differences. The MGA examines differences in relationships across groups rather than interaction effects, thereby identifying group differences in the interpretation of signals. Once partial measurement invariance was achieved using the MICOM procedure, differences in the structural path coefficients were tested using Welch’s t-test.

Since the data were self-reported, procedural and statistical controls were used to minimise common method bias. Harman’s single-factor test revealed no dominant factors. Common method bias was minimised using procedural remedies and assessed using statistical diagnostics, that is, full collinearity VIF, to determine if common method variance was a serious concern for the validity of the results. The instructions in the questionnaire explicitly stated that there were no right or wrong answers, but that respondents should respond honestly. Items representing the different constructs were presented in a logical sequence, and previously validated measurement scales were used to clarify the measurement and minimise item ambiguity.

Table 1 shows that the full collinearity VIF values ranged from 1.004 to 2.848, which were below the conservative threshold of 3.3 recommended by (Kock, 2015) for evaluating potential common method biases. Based on these findings, common method bias is not likely to be a major threat to the validity of the structural relationships.

4. RESULTS

In this study, the PLS-SEM technique was adopted to assess the relationships among omnichannel fulfilment dimensions, perceived supply chain transparency, perceived greenwashing, and customer loyalty for Gen Z and Millennials. The analysis was carried out in two stages: first, the measurement model was tested, and second, the structural model was tested using AMOS. To explore the generational differences between Gen Z and Millennials, the MGA and Measurement Invariance of Composite Models (MICOM) procedures were used. The reported results were derived from the final statistical results presented in the tables.

4.1. Demographic Profile of Respondents

The demographic data of the respondents are presented in Table 2. A total of 410 valid responses were obtained from consumers with omnichannel beauty experience. As is common in the beauty industry, most respondents were women. The sample was reasonably balanced, with both Gen Z and Millennial consumers, allowing multiple comparisons across groups. Most respondents held an undergraduate or postgraduate qualification and purchased beauty products monthly, indicating familiarity with omnichannel fulfilment processes. The overall demographic profile suggested that the sample was suitable for the purpose of the study to explore the perceptions of fulfilment quality dimensions, supply chain transparency, perceived greenwashing, and customer loyalty in sustainable beauty retail.

4.2. Measurement Model Assessment

The indicator reliability, internal consistency, convergent validity, and Average Variance Extracted (AVE) are presented in Table 3 to estimate the measurement model. The reliability of the indicators was satisfactory for both generational groups in the outer loading results. The loadings of most indicators were higher than the recommended value of 0.70, as suggested by (Usakli & Kucukergin, 2018). Within the Gen Z subgroup, a few customer loyalty indicators had loadings just below the general standard of 0.70, which was acceptable. These indicators were retained because the constructs maintained acceptable composite reliability and AVE, and their removal was considered potentially detrimental to content coverage. These indicators would have limited the conceptualisation of customer loyalty to attitudinal and behavioural aspects if they had not been included.

The Cronbach’s alpha values for Gen Z ranged from 0.785 to 0.867, while the composite reliability ranged from 0.874 to 0.899. Likewise, the internal consistency reliability (composite) of the Millennials ranged from 0.872 to 0.922, and Cronbach’s alpha reliability ranged from 0.780 to 0.901 for the Millennials. According to (Sarstedt et al., 2022), the values of the constructs were above the threshold of 0.70, indicating that the overall internal consistency and reliability of the constructs were good. Furthermore, the AVEs computed for all constructs in the Gen Z and Millennial groups exceeded the minimum criterion of 0.50, ranging from 0.595 to 0.748 for Gen Z and 0.658 to 0.725 for Millennials, as detailed by (Mohd Dzin & Lay, 2021). The results indicate the acceptable convergent validity of the measurement model.

To test the discriminant validity, the Heterotrait-Monotrait (HTMT) criterion was used (Table 4). All values of the constructs were below 0.85 (which is the criterion recommended by Yusoff et al., 2020), indicating satisfactory discriminant validity for both generational groups. Considering the somewhat overlapping nature of perceived greenwashing and supply chain transparency, it was important to assess their HTMT values. Empirically, the constructs of perceived greenwashing and perceived supply chain transparency were distinguished with values of 0.109 among Generation Z and 0.079 among Millennials. In general, the results validated the measurement model as reliable and valid for both generational groups.

Table 5 presents the effect sizes (f²) for each generation. The results showed that the effect sizes were small to very low for most relationships, suggesting that individual relationships were modest contributors to endogenous constructs. The largest effect size was observed for perceived supply chain transparency → customer loyalty among Millennials (f² = 0.236), whereas the corresponding effect among Gen Z was smaller (f² = 0.089). Small effects were also observed on Millennials’ perception of supply chain transparency regarding delivery service quality and the efficiency of the delivery process. Overall, the findings indicate that supply chain transparency has the greatest impact on customer loyalty in both generations.

Table 6 reports the positive Q² predict values for the overall sample, representing the two groups. Positive Q² values indicate that the model has a predictive assessment, making it a better predictor of endogenous constructs than simple benchmarks. However, Q² is insufficient for determining the out-of-sample predictive ability. Perceived supply chain transparency had relatively higher predictive relevance (Q² = 0.280), indicating that the model can predict consumer perception of transparency along the supply chain. Likewise, customer loyalty yielded a positive Q² predictive value but was comparatively lower than transparency (Q² = 0.125). Overall, the results suggest that the structural model has good predictive relevance for the constructs of perceived supply chain transparency and customer loyalty in omnichannel retailing.

4.3. Assessment of Explanatory Power

The coefficient of determination (R²) suggests varying degrees of explanatory power for the two generational cohorts (Table 7). The model explained 20.4% of the variance in perceived supply chain transparency and 13.0% of the variance in customer loyalty among Gen Z consumers. However, the model had relatively greater explanatory power for Millennials, explaining 37.2% and 38.0% of the variance in perceived supply chain transparency and customer loyalty, respectively. The results indicate that perceptions of transparency and operational cues related to fulfilment are more explanatory of customer loyalty among Millennials than among Generation Z customers.

4.4. Path Coefficients Assessment

Table 8 presents the extent to which the proposed hypotheses are supported by the structural model. Among Gen Z consumers, delivery service quality was positively and significantly associated with perceived supply chain transparency (p = 0.023; β = 0.260). Likewise, delivery service quality significantly and positively affected Millennials’ perception of supply chain transparency (β = 0.258, t = 2.306, p = 0.021). These results support H1a for both Gen Z and Millennials.

However, H1b was not supported because channel integration was not significantly associated with perceived supply chain transparency in either Generation Z or Millennials. Similarly, Returns Management Efficiency was not significantly associated with perceived supply chain transparency in either cohort. Fulfilment Process Efficiency was not significantly associated with perceived supply chain transparency among Gen Z consumers (β = 0.135, p = 0.285), whereas the association was positive and statistically significant among Millennials (β = 0.277, p = 0.018).

The findings also indicate that both groups had a statistically significant positive structural relationship between perceived supply chain transparency and customer loyalty. There was a significant influence of supply chain transparency on customer loyalty among Gen Z (β = 0.312, t = 3.723, p < 0.001); the Millennial group also showed a statistically significant result for the relationship between transparency and loyalty; however, the estimated coefficient was numerically larger for Millennials (β = 0.484, t = 7.384, p < 0.001). Thus, H2 was supported in both cohorts.

The direct links between fulfilment dimensions and customer loyalty mainly generated non-significant results. For both groups, delivery service quality, channel integration, fulfilment process efficiency, and return management efficiency were not found to have a statistically significant direct effect on loyalty. Thus, H3a–H3d are not supported.

Differences were observed within groups, but no differences were detected between groups using MGA. The interaction between perceived greenwashing and perceived supply chain transparency was negative and statistically significant among Gen Z consumers (β = −0.080, t = 1.980, p < .05), indicating that higher perceived greenwashing was associated with a weaker positive structural association between perceived supply chain transparency and brand loyalty. The effect size was negligible (f² = 0.002). Among Millennials, the interaction was not statistically significant (β = 0.067, t = 0.925, p = .355), with a negligible effect size (f² = 0.006).

The direct effect of perceived greenwashing on the customer loyalty of Gen Z and Millennial consumers was not statistically significant (β = −0.032, t = 0.568, p = 0.570; β = −0.023, t = 0.603, p = 0.546, respectively). These results indicate that the direct association between perceived greenwashing and customer loyalty was not statistically significant in either cohort, whereas its interaction with perceived supply chain transparency was statistically significant only among Gen Z.

4.5. Mediation Results

To test the mediating effect of perceived supply chain transparency between omnichannel fulfilment dimensions and customer loyalty, a mediation analysis was performed (Table 9). The mediation analysis found indirect-only mediation in the case of delivery service quality on customer loyalty and fulfilment process efficiency on customer loyalty in the millennial group. The results indicate that these dimensions of fulfilment have statistically significant indirect associations with customer loyalty through perceived supply chain transparency among Millennials. There were no significant mediation effects in terms of channel integration or return management efficiency for the Gen Z group.

However, for Millennials, perceived supply chain transparency was a statistically significant mediator between delivery service quality and customer loyalty (β = 0.125, t = 2.239, p = 0.025). Transparency also serves as an important mediator between fulfilment process efficiency and customer loyalty (β = 0.134, t = 2.211, p = 0.027). For the remaining fulfilment dimensions i.e., channel integration and returns management efficiency, transparency did not act as a statistically significant mediator/

The findings suggest that among Millennials, some operational fulfilment dimensions are indirectly associated with customer loyalty via perceived supply chain transparency. Accordingly, H4a–H4d were partially supported: H4a and H4c were supported among Millennials, whereas H4b and H4d were not, and none of the four indirect effects were significant for Generation Z.

4.6. MICOM and MGA

As presented in Table 10, most constructs showed full or partial measurement invariance, allowing for a careful comparison between Gen Z and Millennials. However, customer loyalty did not exhibit compositional invariance (p = 0.023). Therefore, the results of the MGA comparisons are not definitive evidence of generational differences in customer loyalty. The fulfilment–loyalty relationship, perceived supply chain transparency–loyalty relationship, and perceived greenwashing moderation of the transparency–loyalty relationship were discussed as exploratory comparisons. Therefore, any lack of a significant difference between groups should not be considered as evidence that the relationships between individuals are the same within these two groups.

After establishing the required level of measurement invariance using MICOM, the Welch–Satterthwaite procedure was used to test the structural path differences between Gen Z and Millennials (Table 11). Because customer loyalty did not establish compositional invariance, comparisons involving this construct were treated as exploratory.

No statistical difference was detected between the structural paths of Gen Z and Millennials in the MGA results. However, because Customer Loyalty did not show compositional invariance, the differences that involved this construct, such as paths from fulfilment to loyalty, transparency of the supply chain to loyalty, and the moderating effect of perceived greenwashing on the transparency-loyalty relationship, were interpreted as exploratory. Thus, the absence of significant group differences in the present sample should be interpreted as a lack of statistical detection of group differences, not as measurement-equivalent structural relationships between the two cohorts. Thus, H6a¹–H6a⁴ were not supported because none of the four fulfilment-to-transparency path differences between Gen Z and Millennials were statistically significant.

The Welch–Satterthwaite comparison indicated that H6b was not supported (p = 0.107). Because customer loyalty was not found to be compositional invariant, this comparison is treated as an exploratory comparison rather than as confirmatory evidence that the transparency–loyalty relationship is the same for both cohorts. Because the moderation path difference was not significant (p = 0.852), H6c was not supported. Since there is no compositional invariance for customer loyalty, the result should be interpreted with caution and should not be interpreted as measurement equivalent moderation between Gen Z and Millennials.

5. DISCUSSION

This study focused on the effect of the operational dimensions of fulfilment as cues that customers can use to form judgments about supply chain transparency in an omnichannel retail context. These findings are consistent with signalling theory and provide an in-depth understanding of the effects of operational fulfilment signals on Gen Z and Millennial transparency perceptions and purchasing behaviour.

H1a–H1d suggested that a positive relationship exists between all fulfilment quality dimensions (i.e., delivery service quality, channel integration, fulfilment process efficiency, and returns management efficiency) and perceived supply chain transparency. H1a was supported, H1b was not supported, H1c was supported for Millennials but not Gen Z, and H1d was not supported. In both cohorts, perceived supply chain transparency was positively related to delivery service quality, with small practical significance. The association between fulfilment process efficiency and transparency was statistically significant for Millennials and had a small effect size, suggesting a limited practical impact. As the MGA did not find a statistically significant difference between cohorts, it should not be assumed that this pattern represents an increase in the importance placed on the efficiency of the fulfilment process for Millennials. Instead, this suggests that this relationship was observed within the Millennial sample. This is also in line with (Peinkofer & Jin, 2023), who posited that the transparency of fulfilment and reliability of operations decrease uncertainty and increase customer perception. (Avvad et al., 2026) proposed that fulfilment competencies are measurable factors for an organisation’s success in an omnichannel retail context. The results show that the transparency of operations and the accountability of the organisation towards customers is timely delivery, service reliability, and service fulfilment. However, channel integration and returns management efficiency did not significantly affect transparency perceptions. This partially challenges the findings of (Yumurtacı Hüseyinoğlu et al., 2018), who found that integration and return processes are significant quality factors in the omnichannel experience. One possible explanation is that channel integration may become a normal part of the omnichannel retail experience, rather than a unique transparency signal. Likewise, returns management may have failed to be an effective transparency signal because the return processes may be linked to extra transportation, packaging, and disposal of the product. However, these explanations were not directly addressed in the present study and should, therefore, be interpreted with caution. Future studies should explore the underlying mechanisms of why these dimensions of fulfilment do not appear to be important transparency signals.

A significant implication of these results is that the dimensions of fulfilment are not necessarily equally informative as operational signals. The dimensions of fulfilment did not exhibit identical patterns of statistical significance across the two samples, although the MGA did not establish statistically significant differences between the cohorts. Delivery service quality was positively and significantly associated with perceived supply chain transparency in both the Gen Z and Millennial samples, with small effect sizes in both groups. However, in this study, the efficiency of channel integration and return management did not act as effective transparency signals. These findings imply that when customers perceive transparency, they tend to focus on high-visibility fulfilment experiences before background operational capabilities.

H2 is supported, indicating a positive association between customer loyalty and supply chain transparency for both groups. This aligns with the findings of (Sansome et al., 2024), who indicated that transparency enhances trust and relational outcomes, and (Gao & Jiang, 2025), who showed a positive impact of transparency on long-term customer engagement in an omnichannel setting. The coefficient for the Millennial cohort was numerically greater than that for the other cohorts, but there was no statistically significant difference between the cohorts. The findings indicate that the perceived transparency of the supply chain positively affects customer loyalty in both generations, and that there is no empirical evidence of generational differences. The results add to transparency research by revealing that customers’ perceived transparency of supply chains is positively linked to their observable fulfilment experiences. This means that when customers make transparency judgments, they may draw on their operational experiences, in addition to what they are told in formal communication, but the current measure does not directly capture the cognitive inference process by which such judgments are made.

Overall, the results did not support H3, which predicted a direct positive relationship between the dimensions of fulfilment and customer loyalty. None of the four dimensions of fulfilment showed a statistically significant direct association with customer loyalty in either cohort within the specified structural models. These results contradict those of previous studies, which have shown that service quality is associated with improved satisfaction and loyalty (Do et al., 2023; Mofokeng, 2023). With sustainability-related omnichannel marketing, the findings suggest that customers may not base their loyalty solely on the fulfilment quality dimensions. However, the mediation results indicate that some dimensions of fulfilment are indirectly associated with customer loyalty through perceived supply chain transparency, particularly among Millennials. The non-significant direct paths indicate that the study did not detect statistically significant direct associations between the four dimensions of fulfilment and customer loyalty in the specified model. However, these results should not be interpreted as evidence that the dimensions of fulfilment do not influence loyalty. The mediation analysis provided more specific evidence for the indirect relationships: perceived supply chain transparency significantly mediated the relationships between delivery service quality and customer loyalty and between fulfilment process efficiency and customer loyalty among Millennials, whereas the corresponding indirect effects were not statistically significant for Gen Z. No statistically significant indirect effects were detected for channel integration or return management efficiency in either cohort.

Accordingly, H4 received partial support because perceived supply chain transparency significantly mediated the relationships between delivery service quality and customer loyalty and between fulfilment process efficiency and customer loyalty among Millennials, whereas the corresponding indirect effects were not statistically significant for Gen Z or for channel integration and return management efficiency. This is in line with signalling theory, which suggests that visible operational signals can influence customers’ overall perception of the organisation (Mollenkopf et al. 2022). However, the present measurement reflects a customer’s perception of transparency, not the cognitive inference process itself; therefore, this interpretation must be interpreted as a theoretical explanation for the observed relationship rather than as a direct test of customers’ inference processes.

Moderation analysis provided limited support for H5. Among Gen Z consumers, the interaction between perceived greenwashing and perceived supply chain transparency was negative and statistically significant (β = −0.080, t = 1.980, p < 0.05), indicating that higher perceived greenwashing was associated with a weaker positive structural association between perceived supply chain transparency and brand loyalty. However, the interaction effect had a negligible effect size (f² = 0.002), indicating a limited practical contribution to the explained variance in customer loyalty. Among Millennials, the interaction was not statistically significant (β = 0.067, t = 0.925, p = 0.355), and the effect size was negligible (f² = 0.006). Thus, the moderation result provided statistical support for H5 only within the Gen Z group, although the interaction effect was small.

Finally, H6a, H6b, and H6c suggested that the interpretation of operational fulfilment signals, perceptions of transparency, and credibility evaluations of greenwashing would be affected by generational cohort membership. The MGA results did not reveal any significant differences between Gen Z and Millennials in any of the proposed structural relationships, although the rationale for analysing cohort-based differences was based on Generational Cohort Theory. These results provide no statistically significant evidence that the tested signalling relationships differ between the two cohorts. In the present study, delivery service quality was positively and significantly associated with perceived supply chain transparency in both cohorts. However, the MGA did not establish a statistically significant difference between the corresponding path coefficients. Overall, no statistically significant between-group differences were detected in the structural paths tested. However, the failure to establish compositional invariance for customer loyalty limits the extent to which cohort comparisons involving this construct can be interpreted. Accordingly, the cohort-specific results are presented as within-group structural associations rather than as evidence that Gen Z and Millennials differ in their underlying relationships.