The state of shopping malls in Jeddah: a consumer-level structural equation modelling approach to behavioural mall resilience in the context of the “death of the mall” debate

1 Introduction

The claim that the shopping mall is in decline has become increasingly prominent over the past decade, driven by accelerating e-commerce penetration, generational shifts in consumption, and post-pandemic restructuring of retail real estate (Pantano and Willems, 2022; Elmashhara and Soares, 2019). However, the “death of the mall” narrative remains analytically contested. Emerging evidence suggests that mall decline reflects lifecycle maturity, competitive sorting, and spatial redistribution rather than universal collapse (Escudero-Gómez and González Aguirre, 2025; Tokosh, 2018; Meijers et al., 2012). Although North American and some European markets report declining footfall, rising vacancies, and the proliferation of dead malls, retail expansion continues across the Gulf Cooperation Council (GCC), particularly in Saudi Arabia. Mall rents in Riyadh increased by approximately 4% year-on-year, with 4.9 million square metres of new retail space planned across major Saudi cities by 2030 (Amawi et al., 2025). At the same time, industry actors report declining footfall and intensifying competition beyond a narrow group of top-performing assets (Upadhyay, 2026). These contradictory trends suggest that the key analytical issue is not whether malls are disappearing, but how mall systems are restructuring under changing economic, spatial, and behavioural conditions (Gilboa et al., 2024).

Recent research increasingly interprets malls as adaptive socio-spatial ecosystems rather than purely transactional retail environments (Gilboa et al., 2024; Roschk and Hosseinpour, 2020). Mall resilience reflects adaptive differentiation and hybrid transformation rather than equilibrium restoration (Teller et al., 2016; Rao and Summers, 2016; Kärrholm et al., 2014). Contemporary malls increasingly integrate retail, leisure, social interaction, and digital consumption into hybrid experiential environments (Saidi et al., 2026; Fares et al., 2026; Mahmud et al., 2023). Mall decline therefore often reflects selective redevelopment, demographic repositioning, and mixed-use urban transformation rather than retail disappearance itself (White et al., 2025; Tokosh, 2018).

These dynamics are highly context dependent. Western “dead mall” trajectories cannot be transferred directly to Saudi Arabia. In the GCC context, socio-cultural norms, climatic conditions, urban accessibility, and institutional arrangements strongly shape retail behaviour and mall patronage (Camrass, 2022; Fernandes and Chamusca, 2014). Their performance is therefore shaped not only by internal retail attributes, but also by broader urban systems including governance, accessibility, surrounding demographics, and spatial restructuring (Tokosh, 2018; Ozuduru et al., 2014; Kärrholm et al., 2014). New retail centralities may erode incumbent malls despite internal adaptation by redistributing accessibility and consumer flows across urban space (Balsas, 2014; Meijers et al., 2012). For this reason, the present study focuses on consumer-level resilience: the behavioural and attitudinal mechanisms through which shoppers continue to visit, prefer, and remain loyal to malls.

Resilience is a multidimensional and multilevel concept whose meaning depends on the system, unit of analysis, disturbance, and outcome being examined. Contemporary resilience research has moved beyond a narrow emphasis on recovery toward a broader understanding involving resistance, adaptation, adjustment, recovery, and transformation across spatial and temporal scales (He et al., 2026). At the urban-system level, resilience concerns the capacity of interdependent ecological, economic, social, institutional, and infrastructural subsystems to maintain essential functions and reorganise under changing conditions. At the retail-district or neighbourhood level, resilience is commonly operationalised through observable structural and spatial characteristics, including retail and functional diversity, accessibility, vacancy, commercial density, socioeconomic capacity, and the ability of retail areas to adapt through successive stages of restructuring (Orr et al., 2023; Suárez et al., 2024; Massarelli and Binetti, 2025; El Baba et al., 2026). A further, complementary perspective concerns the consumer-facing mechanisms through which shoppers continue to visit, prefer, and remain loyal to a retail environment.

The present study adopts this consumer-level perspective. It examines behavioural continuity among current mall users rather than the complete structural, spatial, economic, or institutional resilience of Jeddah’s mall system. Consumer patronage and loyalty may contribute to retail adaptation by sustaining demand, repeated visitation, and relational attachment, but they do not alone establish whether malls are resilient in terms of occupancy, rents, sales, tenant turnover, accessibility, functional diversity, or long-term system transformation. Consumer-level, structural-spatial, and broader system-level approaches are therefore treated as analytically distinct but complementary.

Jeddah provides a critical empirical setting in which these processes can be examined. As Saudi Arabia’s second-largest urban economy, a tourism gateway to the Two Holy Mosques, and a focal point of Vision 2030 retail expansion, the city concentrates both retail growth and competitive pressure. Rather than declining linearly, malls may “bounce forward” through adaptive transformation and experiential repositioning (Barata-Salgueiro and Cachinho, 2021; Barata-Salgueiro and Guimarães, 2020). The central question therefore becomes whether mall loyalty in Jeddah is sustained primarily through transactional utility or through experiential, emotional, and symbolic forms of consumption.

Jeddah has a population of approximately 4.9 million and an estimated 20 major shopping malls supplying around 439,000 m² of gross leasable retail area. Annual retail expenditure in the city is estimated at USD 17–18 billion. Although Riyadh currently leads national retail supply, accounting for 439,000 m² of gross leasable retail area while 1.31 million m² of total space planned in the city by 2030 with a 4.9 million m², Jeddah remains a strategically important retail and tourism hub, reinforced by Vision 2030 giga-projects and religious tourism (Knight, 2025). This concentration of rapid supply growth alongside intensifying inter-mall competition makes Jeddah a critical case for examining consumer-level mall resilience.

Structural indicators such as footfall, vacancy, occupancy, rental performance, tenant turnover, accessibility, and functional diversity are essential for assessing the economic, spatial, and operational resilience of retail systems. Recent studies demonstrate how these indicators can reveal spatial vulnerabilities, adaptive capacity, and patterns of retail restructuring (Orr et al., 2023; El Baba et al., 2026; Fares et al., 2026; Mahmud et al., 2023; Roschk and Hosseinpour, 2020). However, structural indicators do not by themselves explain the perceptual and behavioural mechanisms through which shoppers continue to select, revisit, and remain loyal to a mall. Mall patronage emerges from the interaction of consumer experience, perceived value, attitude formation, and behavioural reinforcement (Gilboa et al., 2024; Ferreira and Paiva, 2017). Empirical evidence further suggests that mall attributes may influence patronage through experiential and affective mechanisms as well as functional and transactional considerations (Fares et al., 2026; Mahmud et al., 2023; Roschk and Hosseinpour, 2020). Accordingly, the present study complements structural-spatial resilience assessment by examining whether loyalty among current shoppers is grounded in functional utility, experiential quality, emotional attachment, or repeated patronage. These consumer-level mechanisms may contribute to retail adaptation, but they do not independently establish whether the wider mall system is economically or spatially resilient.

To examine these mechanisms empirically, this study adopts the Stimulus–Organism–Response (S–O–R) framework (Mehrabian and Russell, 1974), which has been widely revalidated in contemporary retail and atmospherics research (Fares et al., 2026; Kıymalıoğlu et al., 2024; Chen, 2024; Roschk and Hosseinpour, 2020). Within this framework, mall environment, image, merchandise mix, and customer support operate as stimuli; mall experience and attitude represent organismic states through which shoppers interpret mall stimuli; and patronage and loyalty constitute behavioural responses. Hedonic, utilitarian, and emotional value are treated in this study as perceived value appraisals rather than as simple environmental stimuli. This structure enables analysis of how shoppers translate mall attributes and perceived value into behavioural outcomes under conditions of retail restructuring.

The study tests these relationships through a multi-mediator structural equation model (SEM). Experience and attitude function as proximal mediators, while patronage operates as a higher-order mediator linking evaluation to loyalty. The model is applied to consumer data from Jeddah to assess whether these outcomes are sustained primarily through functional utility, experiential quality, emotional attachment, or behavioural reinforcement.

The study addresses three related gaps. First, it provides conceptual clarification by distinguishing consumer-level behavioural resilience from structural-spatial and broader system-level resilience. Recent resilience studies have increasingly adopted multidimensional indicator systems, composite indices, spatial analysis, adaptive-cycle frameworks, coupling-coordination models, and nonlinear explanatory techniques to assess how urban and regional systems resist, adapt, recover, and transform (Suárez et al., 2024; Massarelli and Binetti, 2025; El Baba et al., 2026; He et al., 2026). These approaches are valuable for identifying spatial strengths, vulnerabilities, coordination patterns, and structural conditions, but they do not necessarily explain how individual consumers interpret retail environments and translate those evaluations into continued patronage and loyalty.

Second, the study advances empirical integration by connecting consumer experience, perceived value, attitude, patronage, and loyalty to the wider debate on retail restructuring. Rather than treating patronage and loyalty as complete measures of mall resilience, the study conceptualises them as demand-side behavioural mechanisms that may contribute to the adaptive capacity of physical retail environments.

Third, the study contributes theoretically and methodologically by integrating the S–O–R framework with retail-resilience perspectives through a multi-mediator PLS-SEM model. The theoretical contribution lies in distinguishing external mall stimuli, perceived-value appraisals, organismic states, and behavioural responses while identifying relational, symbolic, affective, experiential, utilitarian, and behavioural pathways to loyalty. The methodological contribution lies in formally testing partial and full mediation and isolating direct heuristic and relational routes. The contextual contribution lies in providing a consumer-level, model-based examination of mall adaptation within a GCC and Vision 2030 setting, while explicitly delimiting the findings from broader structural, spatial, and system-level claims.

The objective of this study is therefore to examine how value drivers influence mall patronage and loyalty in Jeddah, and to determine whether these relationships support, challenge, or refine the death-of-the-mall thesis in Saudi Arabia. More precisely, the study examines whether Jeddah mall loyalty is explained mainly by utilitarian shopping benefits, experiential and hedonic value, emotional attachment, or the conversion of favourable evaluations into repeated patronage.

To achieve this objective, the study addresses the following research questions.

• Which mall attributes most strongly influence shopper experience and attitude in Jeddah?
• Do mall experience and attitude mediate the relationship between mall attributes and patronage?
• Does patronage mediate the relationship between mall experience, attitude, and loyalty?
• Are emotional and hedonic drivers stronger than utilitarian drivers, indicating a shift toward experiential and symbolic consumption?
• Do the findings support a narrative of decline, resilience, or transformation in Jeddah malls?

2 Hypothesis development

2.1 Mall environment, atmospherics and mall experience

Atmospherics including lighting, music, scent, layout, greenery, thermal comfort, and crowding have long been treated as foundational stimulus conditions that shape in-mall experience of the shoppers (Roschk and Hosseinpour, 2020; Turley and Milliman, 2000; Babin et al., 1994). Recent evidence continues to support this premise. In an S–O–R study of factory-outlet malls in Taiwan, atmospherics significantly shaped memorable customer experience, which in turn fostered loyalty (Chen, 2024). Related studies also show that mall aesthetics, biophilic design, multisensory cues, and store atmospherics can strengthen perceived atmosphere, shopper experience, and loyalty-related outcomes (Çavka, 2023; Ameen et al., 2021; Vilnai-Yavetz et al., 2021).

Accordingly, mall environment and atmospherics are specified as external S–O–R stimuli that influence mall experience, which is defined here as the shopper’s overall sensory, cognitive, and affective encounter with the mall during the visit.

Therefore.

H1 Mall Environment and Atmospherics positively influence Mall Experience among shoppers in Jeddah malls.

2.2 Mall image and mall experience

Mall image, the holistic mental representation a shopper holds about a mall, integrating prestige, convenience, and symbolic meaning, has been argued to act as a cognitive frame that shapes on-site experience (El-Adly and Eid, 2016; Sohail, 2013). The image and experience link is grounded in expectancy-confirmation logic; a strong image elevates anticipatory affect, which then colours perceptions of the visit.

Moreover, image is highly endogenous to past experience, raising methodological concerns as a recent study document that direct paths from experience-related antecedents to outcomes can lose significance once mediators are introduced (Pires et al., 2025). This suggests that mall image should be treated carefully as both a prior cognitive frame and a construct that may partly reflect accumulated past experience. In the present model, mall image is specified as an S–O–R stimulus because it represents prior cognitive and symbolic perception of the mall by shoppers, while mall experience refers in-visit encounter of shoppers with the mall. This distinction helps avoid treating image and experience as interchangeable constructs.

Therefore

H2 Mall Image positively influences Mall Experience.

2.3 Hedonic value and mall experience.

Malls act as resource systems that enable identity, leisure, and social life projects for consumers (Cachinho, 2014). Contemporary malls increasingly operate as hybrid consumption ecosystems that integrate retail, leisure, culture, and lifestyle experiences (Barata-Salgueiro and Guimarães, 2020). Research on atmospheric stimuli further shows that sensory retail environments positively shape mood, evaluations, memory, and behavioural intentions, reinforcing the experiential basis of physical shopping environments (Roschk and Hosseinpour, 2020). Shopping has therefore evolved into a socio-spatial experience of identity and leisure, rather than merely transactional consumption (Barata-Salgueiro and Erkip, 2014).

Hedonic shopping value refers to the pleasure, enjoyment, fantasy, and escapism derived from shopping itself (Arnold and Reynolds, 2003; Babin et al., 1994). Because mall experiences are affective and experiential, shoppers who perceive higher hedonic value are likely to evaluate the overall mall experience more positively. In this study, hedonic value is conceptualised as a perceived value appraisal, capturing the extent to which shoppers view mall visits as enjoyable, pleasurable, and recreational. Mall experience, by contrast, refers to the broader sensory, cognitive, and affective encounter with the mall during the visit. Thus, hedonic value is expected to strengthen mall experience by enhancing the emotional and recreational meanings that consumers attach to mall visits.

So

H3 Hedonic Value positively influences Mall Experience.

2.4 Customer support and attitude toward the mall

Customer support refers to the responsiveness, helpfulness, and competence of mall personnel and service systems. In the service-quality literature, supportive and reliable service encounters are commonly treated as important antecedents of customers’ evaluative responses (Brakus et al., 2009; Parasuraman et al., 1988). Within the present S–O–R model, customer support is conceptualised as a service-related stimulus because it reflects the shopper’s interaction with the mall’s human and organisational service environment.

Although self-service technologies and digital concierge tools may reduce dependence on direct staff interaction, they do not eliminate the importance of support quality. Customer support can still influence attitude by shaping perceived care, ease of navigation, problem resolution, and comfort during the visit. Attitude is treated here as the shopper’s overall evaluative disposition toward the mall, distinct from mall experience, which captures the broader in-visit encounter. Therefore, this study proposes that customer support contributes positively to shoppers’ attitudes toward the mall.

So

H4 Customer Support positively influences Attitude toward the Mall.

2.5 Merchandise/tenant mix and attitude toward the mall

Tenant mix, referring to the composition, balance, and complementarity of anchor stores, specialty retailers, food and beverage outlets, and entertainment tenants, has been consistently identified as one of the strongest predictors of mall attractiveness (Xu et al., 2022; Yim and Xu, 2012; Teller, 2008). A well-designed tenant mix enhances the overall shopping environment by offering variety, convenience, and complementary retail and leisure options within a single location. Recent work has also highlighted the increasing importance of tenant compatibility, developing generative-design algorithms to support more effective tenant allocation and retail planning (Wu et al., 2023).

From a consumer perspective, tenant mix functions as a retail-content stimulus that shapes evaluations of the mall’s usefulness, relevance, distinctiveness, and fit with shoppers’ consumption needs. A mall that offers a diverse, yet coherent mix of stores, dining, services, and entertainment is more likely to be perceived as convenient, attractive, and worth visiting. When shoppers perceive the tenant mix as varied, convenient, and aligned with their shopping and leisure purposes, they are more likely to form a favourable attitude toward the mall. Therefore, merchandise and tenant mix are expected to have a positive effect on attitude toward the mall.

Therefore

H5 Merchandise/Tenant Mix positively influences Attitude toward the Mall.

2.6 Emotional value and attitude toward the mall

Emotional value, the affective benefits a shopper accrues from a mall, including social belonging, comfort, attachment, and identity reinforcement, is theorised to translate into evaluative attitudes via affect-as-information mechanisms (Schwarz and Clore, 2003; Sweeney and Soutar, 2001). In this study, emotional value is distinguished from hedonic value: hedonic value refers mainly to enjoyment, pleasure, and recreational gratification, whereas emotional value refers to deeper affective attachment, belonging, security, and identity congruence associated with the mall. It is also distinguished from attitude, which refers to the shopper’s overall evaluative disposition toward the mall.

Prior research suggests that affective responses and emotional appraisals can shape consumers’ overall evaluations of retail environments. Emotional attachment may reflect feelings of comfort, security, belonging, and identity congruity within mall spaces (Gilboa et al., 2024; Voyce, 2003). Similarly, sense of place has been shown to strengthen emotional attachment, patronage, and mall loyalty (Lin et al., 2020). Recent SEM studies also support the role of affective antecedents in shaping evaluative outcomes when emotional-value constructs are carefully separated from attitude constructs (Dang-Van et al., 2023; Han et al., 2021).

Accordingly.

H6 Emotional Value positively influences Attitude toward the Mall.

2.7 Mall experience and patronage

The experience–patronage link is a central proposition of experiential retailing theory (Roschk and Hosseinpour, 2020; Pine and Gilmore, 2011; Verhoef et al., 2009). Recent evidence supports this relationship: memorable customer experience has been shown to mediate the effect of atmospherics on patronage and loyalty (Chen, 2024), while customer experience. Meta-analytic evidence also shows that pleasant atmospheric stimuli can increase customer expenditures and dwell time under favourable experiential conditions (Roschk and Hosseinpour, 2020).

In the present study, mall experience refers to the shopper’s overall sensory, cognitive, and affective encounter with the mall, whereas patronage refers to visit behaviour or intention to visit. A favourable mall experience is therefore expected to increase the likelihood of repeat visitation and continued mall use.

Hence

H7 Mall Experience positively influences Mall Patronage.

2.8 Attitude and patronage

The attitude–behaviour relationship is central to the Theory of Planned Behaviour, where favourable evaluations increase the likelihood of behavioural intention and subsequent action (Ajzen, 1991). This theory is well validated in the context to shopping malls (Yousef, 2025; Wattoo et al., 2025). In this study, attitude refers to the shopper’s overall evaluative disposition toward the mall, while patronage refers to visit behaviour or intention to visit. Therefore, a positive attitude toward the mall is expected to increase mall patronage.

However, this relationship may be weakened by the well-known attitude–behaviour gap, where favourable attitudes do not always translate into actual behaviour because of situational constraints such as traffic congestion, parking limitations, lack of time, online retail alternatives, and competing non-mall leisure venues (Sheeran, 2002).

Therefore

H8 Attitude toward the Mall positively influences Mall Patronage.

2.9 Mall image and patronage

Mall image is theorised to influence patronage directly, in addition to its indirect effect through mall experience. A strong mall image can reduce cognitive effort during choice and activate automatic preferences, consistent with brand-equity reasoning (Sohail, 2013; Keller, 1993).

When mall image carries identity-signalling value through luxury, family-friendly, culturally appropriate, or convenience positioning, it may independently guide patronage as a heuristic choice mechanism under competitive mall conditions (Kaihatu and Spence, 2016).

H9 Mall Image positively influences Mall Patronage directly, over and above its indirect effect through Mall Experience.

2.10 Utilitarian value and patronage.

Utilitarian shopping value refers to the practical benefits consumers obtain from a retail experience, including efficiency, convenience, task accomplishment, product availability, and value-for-money. It has long been associated with patronage among task-oriented shoppers because consumers are more likely to revisit retail environments that help them complete their goals effectively (Babin et al., 1994).

In this study, utilitarian value is treated as a perceived value appraisal rather than as an environmental stimulus. It reflects the extent to which shoppers perceive the mall as useful for completing shopping, dining, service, or leisure-related tasks efficiently. This distinguishes utilitarian value from mall patronage, which refers to visit behaviour or intention to visit.

Prior research supports the role of utilitarian value in shaping consumer loyalty and continued use across retail and digital consumption contexts (Akdim et al., 2022; Evelina et al., 2020).

Hence

H10 Utilitarian Value positively influences Mall Patronage.

2.11 Emotional value and patronage

In addition to its attitudinal route (H6), emotional value is theorised to directly stimulate patronage through approach behaviours predicted by PAD framework (Mehrabian and Russell, 1974), where pleasure and arousal elicit dwelling and revisiting. This proposed direct path tests whether affective attachment can motivate mall visits beyond shoppers’ explicit evaluative attitude. Prior SOR-based research shows that emotional responses can retain direct effects on behavioural intention even after evaluative mediators are introduced. Modern dual-process consumer theory (Pham et al., 2012; Kahneman, 2011) further supports a genuine direct affective route.

So

H11 Emotional Value positively influences Mall Patronage directly, over and above its indirect effect via Attitude.

2.12 Mall experience and loyalty

Customer experience is widely recognised as a decisive antecedent of loyalty, particularly when experiences are memorable, immersive, or emotionally resonant (Chen, 2024; Roschk and Hosseinpour, 2020; Brakus et al., 2009). In this study, mall experience refers to the shopper’s overall sensory, cognitive, and affective encounter with the mall, whereas mall loyalty refers to a stronger commitment expressed through preference, recommendation, resistance to switching, and continued relationship with the mall.

Mall experience is therefore expected to influence loyalty directly by creating memorable, affective, and identity-relevant associations with the mall. Dual-pathway studies further suggest that experience can contribute to loyalty both directly and indirectly through patronage (Chen, 2024).

Therefore

H12 Mall Experience positively influences Mall Loyalty directly, over and above its indirect effect through Patronage.

2.13 Patronage and loyalty

The patronage-loyalty link reflects the classical loyalty cascade (Oliver, 1999); cognitive evaluation produces affective preference, which produces conative intention, which produces action through repeat patronage and finally crystallises into attitudinal loyalty. Behavioural-loyalty literature (Han et al., 2021; Dick and Basu, 1994) confirms that repeated visitation generates familiarity, switching costs, and identity attachment that may produce true attitudinal loyalty rather than mere repeat behaviour. In the present model, patronage refers to visit behaviour or intention to visit, whereas loyalty refers to a deeper commitment reflected in preference, recommendation, resistance to switching, and continued relationship with the mall.

Accordingly

H13 Mall Patronage positively influences Mall Loyalty.

2.14 Customer support and loyalty

Customer support reaches loyalty through both the attitudinal route (H4 → H8 → H13) and a hypothesised direct relational route. Service-recovery and relationship-marketing literature (Grönroos, 2007; Tax et al., 1998) argues that personal service interactions create relational equity that bonds customers to providers beyond transactional satisfaction. In the present model, customer support is treated as a service-related stimulus, and the direct path tests whether support can build relational attachment beyond its indirect effect through attitude and patronage. Recent work also show that customer-support and personnel constructs retain a direct path to loyalty in well-specified SEMs (Fares et al., 2026; Chen, 2024). This supports the inclusion of H14 as a partial-mediation path alongside the indirect route through attitude and patronage.

So

H14 Customer Support positively influences Mall Loyalty directly, over and above its indirect effects via Attitude and Patronage.

2.15 Emotional value and loyalty

The relationship between emotional value and loyalty is well supported by loyalty and affective commitment theory. Treatment of affective commitment, a study on brand experience, and recent S–O–R applications suggest that affective bonds strengthen loyalty by making it more resistant to competitive alternatives, occasional dissatisfaction, and price disadvantages (Fares et al., 2026; Brakus et al., 2009; Oliver, 1999).

In the present model, emotional value refers to the mall’s capacity to generate affective attachment, belonging, comfort, and identity congruence. Mall loyalty, by contrast, refers to preference, recommendation, resistance to switching, and continued relationship with the mall. This distinction is important because emotional value is treated as an affective antecedent rather than as part of the loyalty outcome itself.

Accordingly, H15 is specified as a partial-mediation path. Emotional value is expected to influence mall loyalty directly, while also operating indirectly through attitude and patronage. If the direct path remains significant after attitude and patronage are included, this will indicate that emotional attachment contributes to loyalty beyond positive evaluation and repeat visitation.

So

H15 Emotional Value positively influences Mall Loyalty directly, over and above its indirect effects via Attitude and Patronage.

The model presented in Figure 1 retains five direct paths (H9, H11, H12, H14, and H15) alongside the mediated routes. Partial rather than full mediation is expected on theoretical grounds because each retained direct path corresponds to a distinct mechanism that the proximal mediators, experience, attitude, and patronage, do not fully capture. Mall image (H9) can guide mall choice as a pre-visit heuristic and brand-equity cue before any in-mall experience is formed. Emotional value (H11, H15) can motivate approach and revisiting through an affective route that, under dual-process reasoning, operates in parallel with, rather than only through, deliberate evaluative attitude. Customer support (H14) can create relational equity that bonds shoppers to a mall beyond its effect on attitude and repeat visitation. Mall experience (H12) can leave memorable, identity-relevant traces that shape loyalty directly as well as behaviourally through patronage. Accordingly, the model is specified to allow a formal comparison between full and partial mediation, and the direct paths are treated as theoretically motivated tests rather than as unrestricted additions. The full mediation logic is consolidated in Section 4.

3 Methodology

3.1 Research design and sampling

This study adopts a cross-sectional, quantitative, deductive design to test the proposed S–O–R model among shoppers in Jeddah. A self-administered structured questionnaire was distributed using mall-intercept and online channels across major shopping malls representing diverse positioning tiers (luxury, mid-market, and family-oriented). A purposive non-probability sampling strategy was employed, targeting respondents aged 18 years and above who had visited a Jeddah mall within the preceding 3 months, ensuring recall validity (Hair et al., 2021). Following the 10-times rule and inverse square root method for PLS-SEM minimum sample size (Kock and Hadaya, 2018), a total of 474 valid sample samples were obtained, and the achieved sample exceeded the threshold required to detect path coefficients of |β| ≥ 0.20 at 80% power.

Because the sampling frame targeted recent visitors, it is by construction a current-user population; lapsed users and consumers who have substituted mall visits with e-commerce are not represented. The implications of this design for the interpretation of the death-of-the-mall debate are discussed in the limitations section.

The demographic profile of respondents is summarised in Table 1, including the distribution of respondents across the three mall tiers.

TABLE 1. Respondent profile.

The methodological design reflects the specific level of resilience examined in this study. Recent structural-spatial retail-resilience research has used GIS-based multi-criteria models to combine indicators of diversity, proximity, vacancy, density, and socioeconomic conditions into composite indices, followed by adaptive-cycle classification, hotspot analysis, and sensitivity testing (El Baba et al., 2026). At a broader system level, authors combined a multidimensional indicator system, entropy weighting, a coupling-coordination model, spatial disparity decomposition, and XGBoost-SHAP to examine temporal evolution, spatial heterogeneity, nonlinear associations, and interactions among ecological and economic resilience factors (He et al., 2026).

However, the present study addresses a different analytical question. It uses the individual shopper as the unit of analysis and applies PLS-SEM to estimate relationships among latent perceptions, perceived-value appraisals, experience, attitude, patronage, and loyalty. PLS-SEM is therefore employed to explain how consumer-level behavioural continuity is generated, rather than to construct a spatial or system-level resilience index. GIS and composite-index methods identify where structural resilience is comparatively strong or weak; coupling and machine-learning approaches identify system coordination, temporal patterns, or nonlinear explanatory factors; the present SEM approach identifies how latent consumer evaluations are translated into behavioural outcomes. These approaches are complementary but not interchangeable.

3.2 Measures

All eleven latent constructs were operationalised reflectively using multi-item, five-point Likert scales (1 = strongly disagree; 5 = strongly agree) adapted from validated prior research, including atmospherics and experience (Roschk and Hosseinpour, 2020), mall image (El-Adly and Eid, 2016), hedonic and utilitarian value (Babin et al., 1994), emotional value (Sweeney and Soutar, 2001), customer support (Parasuraman et al., 1988), and loyalty (Oliver, 2014).

Patronage and loyalty are treated as the principal behavioural outcomes of the model and are interpreted as indicators of the consumer-facing component of mall resilience. Patronage captures continued or intended visitation, while loyalty captures a stronger relational commitment expressed through preference, recommendation, resistance to switching, and continued attachment. These constructs do not constitute a comprehensive resilience index and are not treated as substitutes for structural indicators such as vacancy, occupancy, rental performance, sales, tenant turnover, footfall, accessibility, commercial density, functional diversity, or socioeconomic capacity. Nor do they measure wider system dimensions such as economic, ecological, infrastructural, or institutional resilience. The model therefore examines behavioural mechanisms that may contribute to mall adaptation rather than measuring the complete resilience of the retail system.

3.3 Data analysis

Partial Least Squares Structural Equation Modelling (PLS-SEM) was implemented in SmartPLS four because the study is prediction-oriented, includes multiple simultaneous mediators, and examines relationships among latent constructs that cannot be observed directly (Ringle et al., 2024; Hair et al., 2021; Sarstedt et al., 2022). Unlike composite resilience-index approaches (Zhang et al., 2025), which aggregate observable structural or spatial indicators into an overall resilience score, PLS-SEM estimates the relationships among latent consumer perceptions, organismic states, and behavioural outcomes.

The measurement model was evaluated using indicator loadings, Cronbach’s alpha, rho_A, composite reliability, average variance extracted, and the heterotrait-monotrait ratio. The structural model was assessed using path coefficients, coefficients of determination, collinearity diagnostics, and 5,000-sample bootstrapping to derive t-statistics and 95% bias-corrected confidence intervals. Specific indirect effects were assessed to distinguish fully from partial mediation.

The analysis identifies explanatory associations among the study constructs but does not establish temporal or system-level causality. In contrast to longitudinal resilience studies that examine temporal evolution, spatial heterogeneity, subsystem coordination, or nonlinear threshold effects, the present cross-sectional design estimates consumer-level relationships at a single period. Common-method bias was assessed through full-collinearity diagnostics, with all VIF values below 3.3, and Harman’s single-factor test.

4 Synthesis and mediation logic

Together, the fifteen hypotheses specify a multi-stage mediation model in which mall-related stimuli and perceived value appraisals shape consumer-level behavioural outcomes through mall experience, attitude, and patronage. Within the S–O–R framework, mall environment, mall image, merchandise/tenant mix, and customer support are treated as external stimuli, while mall experience and attitude represent organismic states. Hedonic, utilitarian, and emotional value are treated as perceived value appraisals that influence these organismic states and subsequent behavioural responses. Resilience is not specified as a single latent construct in the model. Instead, patronage and loyalty are treated as demand-side behavioural outcomes that may contribute to the adaptive capacity of physical malls.

The model proposes two main mediation routes. First, mall environment, mall image, and hedonic value are expected to influence mall experience, which then influence patronage and, both directly and indirectly through patronage, loyalty. Second, customer support, merchandise/tenant mix, and emotional value are expected to influence attitude, which contributes to patronage and, through patronage, to loyalty. Patronage is positioned as a further mediating mechanism linking both mall experience and attitude to loyalty.

The model also includes theoretically justified direct paths. Five direct paths are retained alongside corresponding mediated routes to test partial rather than full mediation: mall image to patronage, emotional value to patronage, mall experience to loyalty, customer support to loyalty, and emotional value to loyalty. Mall image may guide mall choice as a pre-visit heuristic independently of the subsequent in-mall experience. Emotional value may motivate patronage and loyalty through an affective route that is not fully transmitted through attitude. Mall experience may create memorable and identity-relevant associations that influence loyalty beyond repeated visitation, while customer support may generate relational attachment beyond its effects through attitude and patronage. Utilitarian value is additionally specified as a direct predictor of patronage because it represents the practical usefulness of the mall rather than a parallel mediated route. These paths are included because prior research suggests that cognitive evaluations, functional convenience, emotional attachment, and relational support may influence behavioural outcomes beyond some of the indirect effects transmitted through experience, attitude, and patronage. The model therefore allows comparison between full and partial mediation explanations without treating the retained direct relationships as unrestricted additions.

In the context of Jeddah, this mediation structure helps examine how malls may sustain consumer patronage and loyalty among current mall users amid wider concerns about e-commerce growth and the changing role of physical retail spaces. Strong hedonic and emotional value effects would suggest that malls are functioning increasingly as leisure, social, and identity-related destinations. Strong utilitarian value effects would instead suggest that malls continue to be valued substantially as efficient retail, dining, and service hubs. The relative magnitude and significance of these pathways indicate how effectively favourable perceptions and evaluations are translated into continued visitation and loyalty. Weak experience-to-patronage or attitude-to-patronage relationships would indicate a limited conversion of positive evaluations into behavioural continuity but would not by themselves demonstrate structural or market-wide vulnerability.

The empirical results should therefore be interpreted as evidence of consumer-level behavioural mechanisms that may support mall adaptation, rather than as direct evidence of market-wide mall survival or decline. Patronage and loyalty represent one consumer-facing component of resilience and should not be treated as comprehensive measures of mall-system resilience. Broader claims about structural, spatial, economic, or institutional resilience would require complementary evidence on occupancy, vacancy, footfall, rental and sales performance, tenant turnover, accessibility, functional and tenant diversity, commercial density, surrounding socioeconomic conditions, spatial competition, and changes over time.

5 Results

5.1 Measurement model

All reflective constructs satisfied accepted thresholds for reliability and validity. In general, outer loadings were above 0.70 (Table 2), composite reliability values ranged from 0.831 to 0.870, and Average Variance Extracted values exceeded the recommended 0.50 threshold (Table 3). Variance Inflation Factors remained below 3.3,which, following full collinearity criterion (Kock, 2015), indicates that common-method bias is unlikely to have inflated the structural estimates. As a complementary check, single-factor test (Harman, 1976) showed that the first unrotated factor accounted for only 14.48% of the total variance (below the 50% threshold), providing additional evidence that common-method variance is not a serious concern. Discriminant validity was also established through HTMT analysis (Table 4), confirming that constructs such as Experience, Attitude, Emotional Value, and Hedonic Value were empirically distinct. In particular, the HTMT ratio for the Hedonic Value ↔ Emotional Value pair was 0.167, below the conservative 0.85 threshold.

TABLE 2. Indicator reliability and collinearity.

TABLE 3. Internal consistency reliability and convergent validity.

TABLE 4. Discriminant validity (HTMT).

5.2 Structural model and hypothesis testing

The structural model demonstrated modest but acceptable explanatory power. The model explained 25.6% of variance in Patronage, 25.5% in Attitude, 24.1% in Experience, and 22.6% in Loyalty (Table 5). These findings indicate that consumer-level behavioural mechanisms account for a meaningful but incomplete proportion of mall behaviour in Jeddah.

TABLE 5. Coefficient of determination (R²).

All hypothesised relationships were statistically significant (Table 6). The strongest paths were Customer Support → Attitude (β = 0.315), Image → Experience (β = 0.299), Environment → Experience (β = 0.298), and Tenant Mix → Attitude (β = 0.272). These findings indicate that both experiential and evaluative responses are strongly shaped by service quality, environmental atmospherics, symbolic image, and merchandising strategy.

TABLE 6. Hypothesis testing.

Patronage significantly predicted Loyalty (β = 0.267), supporting the argument that repeated visitation is central to loyalty formation. Image also exerted a substantial direct effect on Patronage (β = 0.246), suggesting that mall image functions as a pre-visit cognitive cue influencing choice independently of in-mall experience. Several weaker but significant effects were also observed. Utilitarian Value retained a direct effect on Patronage (β = 0.157), indicating that task-oriented shopping motivations remain relevant despite the experiential repositioning of contemporary malls. Emotional Value contributed modestly to both Patronage and Loyalty, while Hedonic Value had a weaker-than-expected effect on Experience (β = 0.143). This suggests that environmental and symbolic dimensions of the mall experience may outweigh purely recreational motivations in the Jeddah context.

5.3 Mediation effects

The results support partial rather than full mediation. Image and Customer Support retained strong direct effects on Patronage and Loyalty respectively, indicating that symbolic and relational mechanisms influence consumer behaviour independently of experiential evaluations. Experience and Attitude functioned as bounded mediators, with their effects on Loyalty operating primarily through Patronage. This supports the view that loyalty develops through behavioural reinforcement and repeated visitation rather than through isolated positive experiences alone. The specific indirect effects are reported in Table 7. Comparing the two proximal mediators, the attitude-mediated route and the experience-mediated route to loyalty (via patronage) can be evaluated directly from their respective indirect coefficients and confidence intervals.

TABLE 7. Specific indirect (mediated) effects.

Five key inferences emerge from the empirical model. First, Customer Support and Mall Image are the strongest predictors of mall behaviour in Jeddah. Second, mall experiences are shaped primarily by environmental and symbolic factors rather than by hedonic value alone. Third, loyalty is driven mainly through repeated patronage and relational service quality. Fourth, emotional value contributes to behavioural outcomes but with comparatively modest magnitude. Finally, utilitarian shopping motives continue to influence mall patronage despite broader experiential trends in retail development.

6 Discussion

The findings of this study contribute to contemporary debates on consumer-level mall resilience, experiential retail transformation, and the contextual applicability of the S–O–R framework within non-Western retail environments, particularly in rapidly transforming Gulf cities such as Jeddah. The results collectively suggest that the death-of-the-mall thesis, which has dominated much of the North American and European retail literature (Pantano and Willems, 2022; Tokosh, 2018), cannot be straightforwardly generalised to the Saudi context. Instead, the evidence points toward a process of selective restructuring and continued behavioural relevance among current mall users, sustained through relational, symbolic, environmental, and experiential affective, and utilitarian mechanisms However, because the analysis does not incorporate structural indicators or lapsed users and non-users, these conclusions explain why current shoppers remain attached to malls rather than demonstrating market-wide survival, structural adaptation, or long-term economic resilience.

This interpretation reflects a multilevel understanding of resilience. The present study examines one demand-side component of retail adaptation: the mechanisms through which current shoppers translate mall attributes and perceived value into experience, attitude, patronage, and loyalty. Structural-spatial resilience research, by contrast, examines conditions such as accessibility, vacancy, commercial density, functional diversity, socioeconomic capacity, and spatial variation, while broader system-level research assesses resistance, adaptation, recovery, transformation, temporal change, and coordination among interdependent subsystems (El Baba et al., 2026; He et al., 2026). These perspectives are complementary rather than interchangeable. Consumer attachment may support mall adaptation, but it does not independently establish that the wider retail system is spatially, economically, institutionally, or operationally resilient.

A central implication of the findings is that malls in Jeddah continue to possess consumer-level competitive advantages that digital retail platforms may find difficult to replicate. Although the present study does not directly measure structural indicators such as vacancy rates, rental decline, sales performance, occupancy, tenant turnover, or footfall, the PLS structural model demonstrates that Customer Support, Mall Image, and Mall Environment significantly vshape consumer evaluations and behavioural outcomes. This finding reinforces arguments advanced in contemporary retail and experiential-retailing literature that malls increasingly function not merely as transactional marketplaces, but as socio-spatial ecosystems that facilitate identity expression, leisure consumption, social interaction, and experiential engagement (Gilboa et al., 2024; Roschk and Hosseinpour, 2020). In the Saudi context specifically, climate-controlled environments, family-oriented leisure practices, and culturally embedded patterns of social gathering may help explain continued attachment to physical mall spaces. Because these contextual conditions were not measured directly, however, their explanatory role should be interpreted as a context-based proposition rather than as an independently tested empirical finding.

Nevertheless, the findings also caution against an overly optimistic interpretation of consumer-level mall resilience. The relatively modest explanatory power of the model (R² values ranging from 0.226 to 0.256) indicates that the included constructs explain only a limited proportion of variation in experience, attitude, patronage, and loyalty. The unexplained variance is likely to reflect determinants not captured in the present consumer-level model, including accessibility and travel time, parking availability, household income, individual online-shopping propensity; mall-specific differences, competing leisure alternatives, and wider neighbourhood conditions; incorporating these factors is a priority for future work. Structural pressures identified in the wider literature, including e-commerce expansion, changing urban retail centralities, evolving mobility patterns, and generational transformations in shopping behaviour, remain potentially destabilising forces for long-term mall sustainability (Elmashhara and Soares, 2019; Balsas, 2014; Meijers et al., 2012).

Recent resilience research further suggests that favourable performance in one dimension does not automatically generate resilience across the wider system. Consumer loyalty may coexist with declining occupancy, spatial inaccessibility, weak tenant diversity, or deteriorating commercial performance. Similarly, structural strength in one period may not ensure continued adaptation when technological, economic, or spatial conditions change. Comprehensive assessment therefore requires attention to multiple dimensions, spatial heterogeneity, temporal trajectories, and potentially nonlinear or threshold relationships (El Baba et al., 2026; He et al., 2026).

The continued significance of Utilitarian Value as a predictor of patronage further suggests that a substantial proportion of mall visits remain functionally and transactionally motivated. Such visits may be comparatively more vulnerable to substitution by online retail systems, particularly under conditions of increasing digital integration and logistical efficiency. Consequently, the findings support a narrative of selective and consumer-level resilience rather than unconditional mall-system stability.

Among all explanatory mechanisms, Customer Support emerged as one of the most influential mechanisms in the model. Its substantial influence on Attitude (β = 0.315) and its strong direct effect on Loyalty (β = 0.230) demonstrate that relational service quality remains a critical source of competitive differentiation for physical retail environments. This finding aligns with relationship-marketing and service-quality theory, which emphasise the continuing importance of interpersonal interaction, trust formation, and perceived care in shaping consumer attachment (Grönroos, 2007; Parasuraman et al., 1988). In the context of Saudi Arabia, the significance of Customer Support may also reflect broader socio-cultural norms surrounding hospitality, attentiveness, and personalised service expectations. Unlike purely digital retail environments, malls provide opportunities for human interaction, reassurance, and social responsiveness that may strengthen emotional and behavioural loyalty. The direct Customer Support–Loyalty relationship also indicates that relational attachment is not transmitted exclusively through general attitude and repeated patronage, supporting the model’s partial-mediation specification.

Mall Image also emerged as a particularly influential mechanism within the PLS structural model. Notably, the direct effect of Mall Image on Patronage exceeded its indirect effect mediated through Experience, suggesting that image operates substantially as a heuristic and symbolic choice mechanism rather than solely as an experiential enhancer. Consumers appear to rely on perceptions of prestige, symbolic reputation, architectural identity, and tenant status when selecting malls before the actual visit experience occurs. This finding is theoretically important because it extends branding and retail-image literature by demonstrating that symbolic positioning functions as a pre-consumption behavioural filter under competitive retail conditions (Sohail, 2013; Keller, 1993). In highly competitive urban retail systems such as Jeddah, malls therefore compete not only through functional offerings but also through symbolic differentiation and status signalling. This pathway represents a consumer choice mechanism and should not be interpreted as direct evidence that a strong mall image produces structural or financial resilience.

Another theoretically significant outcome concerns the relatively weak influence of Hedonic Value on Mall Experience. Contemporary experiential-retail literature frequently argues that malls increasingly compete through entertainment, leisure, and recreational consumption (Barata-Salgueiro and Guimarães, 2020; Pine and Gilmore, 2011). However, the comparatively modest effect observed in this study suggests that purely recreational motivations may be less behaviourally decisive than environmental atmospherics and symbolic image in the Jeddah context. One possible explanation is that malls in Saudi cities have historically functioned as multifunctional socio-cultural environments rather than exclusively commercial destinations. Consequently, entertainment and leisure dimensions may no longer generate the same novelty sometimes attributed to experiential retail development. Instead, shoppers may perceive leisure integration as a normalised baseline expectation rather than as a distinctive competitive advantage. This explanation remains interpretive, however, because the study does not directly compare Saudi and Western shoppers or measure changes in the perceived novelty of mall-based leisure.

The findings further support the classical “loyalty cascade” proposed in loyalty theory (Oliver, 2014). Experience and Attitude influenced Loyalty primarily through Patronage, indicating that loyalty formation depends substantially on repeated behavioural reinforcement rather than isolated positive evaluations. In practical and theoretical terms, this suggests that memorable experiences alone may not be sufficient for sustaining long-term loyalty unless they are translated into recurring visitation behaviour. The results therefore reinforce behavioural-loyalty perspectives arguing that familiarity, routine reinforcement, and repeated interaction are central to durable consumer attachment (Dick and Basu, 1994). Mall loyalty in Jeddah appears to emerge incrementally through sustained patterns of visitation rather than through singular experiential encounters. At the same time, the retained direct Experience–Loyalty path indicates that memorable or identity-relevant experiences may also leave an effect beyond behavioural repetition, supporting partial rather than complete mediation.

The study contributes to the literature in four related ways. First, contextually, it extends S–O–R modelling into the GCC retail context, where empirically grounded research on mall behaviour remains comparatively limited despite rapid retail urbanisation. Second, conceptually, it distinguishes consumer-level behavioural resilience from structural-spatial and broader system-level resilience. Patronage and loyalty are treated as demand-side mechanisms that may contribute to adaptation, rather than as comprehensive indicators of market-wide survival. Third, theoretically, the results demonstrate that continued consumer attachment operates through multiple parallel pathways—relational, symbolic, experiential, affective, utilitarian, and behavioural—rather than through a single experiential route. Fourth, methodologically, the study specifies a multi-mediator PLS-SEM that formally distinguishes partial from full mediation and isolates direct heuristic, affective, experiential, and relational pathways, including Mall Image → Patronage, Emotional Value → Patronage and Loyalty, Mall Experience → Loyalty, and Customer Support → Loyalty.

These contributions also help reposition the “death-of-the-mall” debate away from a simple binary distinction between survival and decline. The findings do not demonstrate that Jeddah’s malls are structurally or economically secure; rather, they identify the forms of consumer attachment that remain viable among current users under conditions of retail transformation. The more appropriate question is therefore not simply whether malls are alive or dead, but whether consumer attachment, structural performance, spatial accessibility, tenant diversity, and institutional adaptation reinforce one another over time. Answering that broader question will require integrating consumer-level modelling with structural indicators, geospatial analysis, and longitudinal evidence.

7 Conclusion and recommendations

The findings support a cautiously optimistic interpretation of the consumer-facing dimension of mall resilience in Jeddah at the consumer level. Physical malls continue to retain behavioural relevance among current users through relational service quality, symbolic image, environmental atmospherics, perceived value, repeted patronage, and loyalty (Chen, 2024; Gilboa et al., 2024; Roschk and Hosseinpour, 2020). These dimensions are less readily reproduced through exclusively digital retail channels and may therefore remain important sources of competitive advantage for physical retail environments (Fares et al., 2026; Ameen et al., 2021; Verhoef et al., 2009). The findings nevertheless demonstrate consumer attachment rather than comprehensive structural, spatial, economic, or institutional resilience.

Continued consumer attachment should not be assumed to translate automatically into the long-term resilience of the wider mall system (Barata-Salgueiro and Cachinho, 2021; Teller et al., 2016). Mall sustainability increasingly depends upon cultivating relational attachment, symbolic differentiation, experiential environments, and consumer engagement (Mahmud et al., 2023; Barata-Salgueiro and Guimarães, 2020; Grönroos, 2007; Tax et al., 1998). Mall managers should therefore supplement transactional retail function with experiential, relational, and symbolic strategies that convert favourable evaluations into repeated visits and loyalty under intensifying competition (Saidi et al., 2026; Pantano and Willems, 2022; Elmashhara and Soares, 2019).

More specifically, the results indicate that customer support should be treated as a strategic capability rather than a secondary service feature. Investment in staff responsiveness, problem resolution, wayfinding assistance, digital–human service integration, and culturally appropriate hospitality may strengthen both attitude and loyalty. Mall image should likewise be managed through coherent architectural identity, positioning, tenant composition, communication, and symbolic differentiation. Environmental quality, including thermal comfort, layout, cleanliness, lighting, ambience, and opportunities for social interaction, should be maintained as a core component of the in-mall experience. At the same time, the continuing significance of utilitarian value indicates that experiential development should not come at the expense of convenience, product availability, service accessibility, task efficiency, and value for money.

The managerial implication is therefore not that malls should transform exclusively into entertainment destinations. Rather, resilient consumer attachment appears to depend on the integration of relational service, symbolic positioning, environmental quality, practical utility, and opportunities for repeated engagement. Because patronage mediates much of the relationship between experience, attitude, and loyalty, managers should also focus on converting favourable evaluations into recurring visits through relevant tenant mixes, event programming, loyalty initiatives, personalised communication, and consistent service delivery.

The following planning implications extend beyond the constructs measured in this consumer-level study and are therefore offered as tentative, context-based suggestions rather than direct empirical findings. At the planning level, authorities should evaluate proposed retail development in relation to existing commercial capacity, accessibility, surrounding land uses, competing retail centralities, and neighbourhood needs rather than promoting indiscriminate retail expansion (Tokosh, 2018; Kärrholm et al., 2014; Meijers et al., 2012). Retail accessibility should be integrated into broader mobility and transport-planning frameworks (Balsas, 2014; Ozuduru et al., 2014), while mixed-use redevelopment and reuse strategies should be considered for retail assets as conditions evolve under Vision 2030 (Amawi et al., 2025; White et al., 2025; Escudero-Gómez and González Aguirre, 2025). Planning assessment should therefore extend beyond consumer perceptions to include public-transport connectivity, pedestrian access, parking conditions, travel time, functional and tenant diversity, vacancy, commercial density, catchment demographics, and relationships with surrounding commercial streets. Recent structural-spatial research demonstrates that retail resilience varies locally and is shaped by diversity, proximity, walkability, socioeconomic conditions, and spatial context (Orr et al., 2023; El Baba et al., 2026). Urban authorities could use these indicators to identify areas of overprovision, vulnerability, spatial exclusion, or redevelopment potential.

Planning strategies should also recognise that consumer, structural, spatial, infrastructural, and economic dimensions of resilience may not evolve uniformly. A mall may retain loyal shoppers while experiencing rising vacancy, declining rental performance, limited accessibility, or weakening tenant diversity. Conversely, a structurally well-located mall may fail to create meaningful consumer attachment. Resilience interventions should therefore be place-specific and should account for developmental stage, local competition, mobility conditions, neighbourhood characteristics, and potential nonlinear or threshold effects rather than assuming that a uniform policy will produce comparable outcomes across all malls and catchments (El Baba et al., 2026; He et al., 2026).

Where existing malls show signs of structural vulnerability, adaptive reuse and mixed-use transformation may be more appropriate than continued retail expansion. Potential interventions include introducing community services, cultural facilities, flexible workspaces, public amenities, residential components, health services, and complementary leisure functions, provided that such strategies are supported by local demand, accessibility, and feasibility assessments. These recommendations remain planning propositions and were not tested directly by the present consumer survey.

Several limitations should be borne in mind when interpreting these findings. First, the study measures consumer-level perceptions, patronage, and loyalty; it does not incorporate structural indicators of resilience such as occupancy rates, rental performance, sales trends, tenant turnover, or footfall. It also does not measure vacancy, accessibility, commercial density, functional diversity, spatial competition, or neighbourhood socioeconomic conditions. The findings are therefore restricted to consumer-level behavioural mechanisms that may contribute to adaptation and should not be interpreted as direct evidence of market-wide survival, structural transformation, or economic resilience.

Second, the sample was recruited through mall-intercept and online surveys of respondents who had visited a mall within the previous 3 months. By construction it captures current mall users and excludes lapsed users and consumers who have substituted mall visits with e-commerce; the results consequently explain the behaviour of active shoppers rather than the whole consumer market and qualify rather than refute the death-of-the-mall thesis. The sample may also underrepresent individuals who avoid malls because of travel time, accessibility barriers, cost, limited relevance, parking difficulties, or preference for alternative leisure and retail environments.

Third, the cross-sectional design precludes strong causal and temporal claims about the hypothesised sequence experience → attitude → patronage → loyalty; longitudinal panel designs and experimental (for example, scenario-based) studies are needed to establish temporal precedence. Resilience is inherently dynamic and involves resistance, adaptation, recovery, and transformation over time. A single-period survey cannot determine whether the observed consumer relationships persist during economic disruption, technological change, new retail competition, transport-system change, or shifts in household behaviour.

Fourth, common-method bias was addressed procedurally and through full collinearity and Harman’s single-factor diagnostics, but a single-source, self-report design remains a limitation; future work could adopt marker-variable or multi-source designs. Self-reported patronage intentions may not correspond fully with observed visit frequency, dwell time, expenditure, recommendation behaviour, or switching. The model would therefore benefit from validation using transaction records, loyalty-programme data, mobile-location information, footfall counts, or other behavioural measures, subject to appropriate privacy and ethical safeguards.

Fifth, the modest explanatory power of the endogenous constructs indicates that important determinants remain outside the present model. These may include accessibility, travel time, parking, household income, household composition, mall distance, competing leisure opportunities, online-shopping propensity, digital-service quality, perceived safety, and mall-specific characteristics.

Sixth, the analysis does not capture the hierarchical and spatial structure of the data. Individual shoppers are nested within malls, catchment areas, neighbourhoods, and wider urban mobility systems. Differences among luxury, mid-market, and family-oriented malls may influence the strength of the estimated relationships, but the present design does not model these contextual effects explicitly. Multilevel and spatially explicit approaches would be required to separate individual-level relationships from mall- and neighbourhood-level influences.

Finally, PLS-SEM is appropriate for testing theoretically specified direct and mediated relationships among latent constructs, but it does not independently reveal complex nonlinearities, threshold effects, unobserved consumer segments, or spatially heterogeneous pathways. Complementary segmentation, multigroup, finite-mixture, spatial, and interpretable machine-learning approaches may provide additional insights, although such methods should be theoretically justified and validated on suitable data.

Future research should integrate consumer-level behavioural modelling with structural retail indicators such as footfall, vacancy rates, rental trajectories, sales density, and accessibility patterns (Upadhyay, 2026; Tokosh, 2018; Ozuduru et al., 2014). This integration could be developed through a multilevel framework in which individual shoppers are nested within malls, mall catchments, and neighbourhood retail systems. At the individual level, S–O–R constructs could explain experience, attitude, patronage, and loyalty. At the mall or catchment level, these relationships could be examined alongside occupancy, vacancy, tenant turnover, tenant and functional diversity, footfall, rental trajectories, sales performance, commercial density, accessibility, transport connectivity, mixed-use intensity, and surrounding socioeconomic conditions.

A complementary structural retail-resilience index could also be constructed for individual malls or catchment areas. This would permit direct examination of whether stronger consumer experience, patronage, and loyalty correspond with lower vacancy, greater tenant retention, more stable footfall, stronger sales performance, or better accessibility. Such analysis would clarify whether behavioural and structural dimensions reinforce one another or diverge under particular market and spatial conditions.

Comparative research across GCC cities and longitudinal analyses would clarify how mall resilience evolves under changing demographic, technological, and economic conditions (Gilboa et al., 2024; Camrass, 2022; Fernandes and Chamusca, 2014). Such research would help bridge the gap between behavioural resilience and structural retail sustainability (Rao and Summers, 2016; Teller et al., 2016).

Consequently, the findings suggest that malls in Jeddah are not experiencing uniform decline. Rather, they are undergoing functional restructuring in which relational, symbolic, experiential, and behavioural dimensions increasingly determine consumer-level resilience (Fares et al., 2026; Gilboa et al., 2024; Barata-Salgueiro and Guimarães, 2020). The “death-of-the-mall” narrative therefore appears insufficient, on its own, for explaining the contemporary Saudi retail landscape (Escudero-Gómez and González Aguirre, 2025; White et al., 2025). A more accurate interpretation is that malls are evolving from transactional retail infrastructures into adaptive socio-spatial environments whose viability depends on sustaining consumer attachment under retail transformation (Saidi et al., 2026; Barata-Salgueiro and Cachinho, 2021; Kärrholm et al., 2014; Cachinho, 2014).

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. Authors acknowledge the use of ChatGPT 5.5 solely in enhancing the clarity and language of this work. The outputs generated were carefully reviewed and edited by the authors, who accepts full responsibility for the content of the publication.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Philip Boland, Queen’s University Belfast, United Kingdom

Reviewed by: Minzhe Du, South China Normal University, China

Nazish Abid, University of Bahrain, Bahrain