Knowledge sharing and innovation performance: a case study on the impact of organizational culture, structural capital, human resource management practices, and relational capital of real estate agents

The paper examines how organizational and individual factors shape innovation performance among real estate agents in Taiwan. Innovation performance reflects an organization’s ability to convert innovation inputs into outcomes and is critical across sectors. Prior real estate studies have emphasized financial and service metrics; this study extends the lens to knowledge processes and intangibles. The authors posit that organizational culture, structural capital, and HRM practices at the organizational level, together with relational capital and knowledge sharing at the individual level, influence agents’ innovation performance. Given the service-oriented, commission-based, and high-turnover nature of Taiwan’s real estate industry, understanding the role of knowledge sharing and contextual factors is both practically and theoretically important. The study tests whether HRM motivates knowledge sharing and innovation, whether structural capital (intangible organizational assets and processes) builds competitive advantages, and whether relational capital (trust, commitment, information exchange) enables knowledge sharing and innovation.

The literature links innovation performance to internal culture, knowledge processes, and social/structural capital. Organizational culture (customer/competitor orientation and interdepartmental coordination) supports innovation (Daft, 2004; Shahzad et al., 2017) and knowledge sharing (Tushman & O’Reilly, 1996; McDermott & O’Dell, 2001). Structural capital—organizational, process, and innovation capital—enhances value creation and innovation (Edvinsson & Malone, 1997; De Pablos, 2004). Relational capital—trust, commitment, and information exchange—supports knowledge integration and innovation (Nonaka & Takeuchi, 1995; Tu, 2009; Onofrei et al., 2020). Knowledge sharing consistently predicts innovation performance (Lin, 2007; Wang & Hu, 2020). HRM practices can enhance innovation, potentially moderating the knowledge sharing–innovation link (Papa et al., 2020; Kim & Park, 2017). The study formulates 12 hypotheses: H1–H3 (organizational culture, structural capital, relational capital → innovation performance), H4 (knowledge sharing → innovation performance), H5–H7 (organizational culture, structural capital, relational capital → knowledge sharing), H8 (HRM practices → innovation performance), H9 (HRM moderates knowledge sharing → innovation performance), and H10–H12 (knowledge sharing mediates the effects of organizational culture, structural capital, and relational capital on innovation performance). The review also notes industry digitization via PropTech, creating new contexts for knowledge sharing and innovation.

Design: A multilevel, cross-sectional survey analyzed with hierarchical linear modeling (HLM) to address nesting of agents (Level 1) within branches (Level 2) and to test mediation and moderation paths. Sampling and data collection: Convenience sampling of seven real estate chains in Kaohsiung City (Sinyi, HandB, Taiching, Taiwan Realty, Yung Ching, CTBC Real Estate, U-Trust). Surveyed commercial hubs in Sanmin, Zuoying, Lingya, Gushan, and Xinxing districts. Administration: in-person before mid-May 2021 and by mail afterward due to COVID-19. Field period: May 1–July 31, 2021. Distributed 1,130 questionnaires; 444 returned; 40 invalid (missing items or no sex/tenure), plus 3 removed where branches had <3 responses. Final N=401 valid respondents from 55 branches (effective response rate 35.49%). Nonresponse bias checks (Armstrong & Overton, 1977) across sex, marital status, education showed no significant differences between early (in-person) and later (mail) responders. Measures: All items measured on 5-point Likert scales (1=strongly disagree to 5=strongly agree). Organizational-level (Level 2) constructs: organizational culture (artifacts, espoused values, basic assumptions; 9 items; α=0.949), structural capital (organizational, innovation, process capital; 8 items; α=0.962), HRM practices (HR planning; training and development; remuneration and benefits; 8 items; α=0.953). Individual-level (Level 1) constructs: relational capital (mutual trust, commitment, information exchange; 8 items; α=0.968), knowledge sharing (sharer, recipient, intentions; 8 items; α=0.960), and innovation performance (stimulating innovation, service innovation; 5 items; α=0.956). Item sources are documented; all factor loadings were significant, indicating construct validity; convergent validity exceeded recommended thresholds (loadings >0.5). Controls: Sex, tenure (job tenure), and business model (direct sales vs. franchise) included as controls. Aggregation and reliability: Level 2 variables were created by aggregating individual responses to branch-level means. Justification via ICC(1) and ICC(2): ICC(1) for organizational culture and structural capital=0.998; ICC(2) for organizational culture, structural capital, and HRM practices≈0.999, indicating strong between-group reliability. Within-group agreement r_wg(j): organizational culture=0.973, structural capital=0.966, HRM practices=0.950 (>0.70), supporting aggregation. Analytical strategy: HLM with mediation configurations 1→1→1 and 2→1→1 (Krull & MacKinnon, 1999). Null models assessed between-branch variance and ICCs: Innovation performance random intercept variance τ00=0.086 (SE=0.293), residual σ2=0.328; ICC=0.208; Knowledge sharing τ00=0.068 (SE=0.261), σ2=0.292; ICC=0.189; both p<0.001, justifying HLM. Mediation tested via Baron & Kenny (1986) three-step approach across: (1) organizational culture→knowledge sharing→innovation performance; (2) structural capital→knowledge sharing→innovation performance; (3) relational capital→knowledge sharing→innovation performance. Moderation: cross-level interaction of HRM practices on the knowledge sharing→innovation performance slope. Models estimated random intercepts (Level 1), fixed slopes (unless specified). Deviance statistics reported for model fit. Estimation outputs: Key fixed effects and random components reported in Table 3 for Model 1 (DV=innovation performance), Model 2 (DV=knowledge sharing), and Model 3 (DV=innovation performance with mediator and moderator).

- Null models: Significant between-branch variance for innovation performance (τ00=0.086, p<0.001; ICC=0.208) and knowledge sharing (τ00=0.068, p<0.001; ICC=0.189), supporting multilevel analysis. - Model 1 (DV: innovation performance): • Relational capital: β=0.643, SE=0.050, p<0.01. • Organizational culture (Level 2): γ=0.556, SE=0.154, p<0.01. • Structural capital (Level 2): γ=0.381, SE=0.176, p<0.05. • Controls: Sex β=0.125, SE=0.045, p<0.01; Tenure β=0.002, n.s.; Business model n.s. • Deviance (-2LL)=496.270; parameters=11. - Model 2 (DV: knowledge sharing): • Organizational culture: γ=0.327, SE=0.132, p<0.05. • Structural capital: γ=0.504, SE=0.158, p<0.01. • Relational capital: β=0.691, SE=0.043, p<0.01. • Controls: Sex β=0.067, SE=0.039, p<0.05; Tenure β≈0.009, p<0.05. • Deviance (-2LL)=381.350; parameters=11. - Model 3 (DV: innovation performance; includes mediator and moderation): • Knowledge sharing: β=0.580, SE=0.062, p<0.01. • Organizational culture: γ=0.605, SE=0.124, p<0.01. • Structural capital: γ=0.004, SE=0.208, n.s. (direct effect becomes non-significant when knowledge sharing included). • Relational capital: β=0.250, SE=0.067, p<0.01. • HRM practices (direct effect): γ=0.317, SE=0.147, p<0.05. • Moderation (HRM practices × knowledge sharing slope): γ=−0.048, SE=0.137, n.s. • Controls: Sex β=0.093, SE=0.039, p<0.01; Tenure β≈0.009, p<0.01. • Deviance (-2LL)=355.169; parameters=16. - Mediation conclusions: • H10 supported: Knowledge sharing partially mediates organizational culture→innovation performance (organizational culture remains significant alongside knowledge sharing). • H11 supported: Knowledge sharing fully mediates structural capital→innovation performance (direct effect non-significant with mediator present). • H12 supported: Knowledge sharing partially mediates relational capital→innovation performance (both relational capital and knowledge sharing significant). - Hypotheses summary: Supported: H1, H3–H8, H10–H12. Not supported: H2 (direct), H9 (moderation).

Findings demonstrate that knowledge processes bridge organizational context and individual innovation outcomes in real estate agencies. Organizational culture enhances knowledge sharing, which, in turn, drives higher innovation performance; culture also exerts a direct positive effect, yielding partial mediation. Structural capital’s contribution to innovation operates through knowledge sharing (complete mediation), suggesting that systems, routines, and innovation assets must be translated into actual knowledge exchanges to affect performance in this context. Relational capital (trust, commitment, information exchange among colleagues) both directly and indirectly contributes to innovation performance, emphasizing the importance of strong interpersonal ties for sharing and applying knowledge. HRM practices positively and directly affect innovation performance but do not moderate the knowledge sharing–innovation relationship, perhaps reflecting limited emphasis or uneven perception of HRM mechanisms in commission-based, high-autonomy agency settings. Collectively, the results support a multilevel model where organizational-level culture and structural capital enable knowledge sharing, and individual-level relational capital activates exchange, culminating in improved innovation outcomes for agents. The results are consistent with broader literature linking culture and social/structural capital to knowledge sharing and innovation, while highlighting sector-specific dynamics that may attenuate direct structural capital effects absent active knowledge sharing.

The study contributes a multilevel framework linking organizational culture, structural capital, HRM practices (organizational level), and relational capital and knowledge sharing (individual level) to innovation performance among real estate agents. Using HLM on 401 agents across 55 branches, the authors show that (1) organizational culture directly and indirectly (via knowledge sharing) enhances innovation performance; (2) structural capital influences innovation only through knowledge sharing (full mediation); (3) relational capital affects innovation both directly and via knowledge sharing (partial mediation); and (4) HRM practices have a positive direct effect but do not moderate the knowledge sharing–innovation link. Managerially, strengthening a cooperative, fair, and engaging culture; investing in systems, training, and process support that promote knowledge exchange; fostering trustful relationships; and designing clear HRM policies (career paths, training application, equitable remuneration) can enhance innovation performance. Future research could broaden HRM dimensions (e.g., performance appraisal, non-financial rewards), include supervisors’ perspectives and job position effects, extend beyond Kaohsiung to improve generalizability, and examine technological factors (e.g., AI/PropTech) and buyer behavior, including potential moderating effects of AI on knowledge sharing–innovation pathways.

- Geographic and sample scope: Data limited to real estate agents in Kaohsiung City, Taiwan; results may not generalize to other regions or sectors. - Control variables: Only sex, tenure, and business model included; job position excluded and supervisors not surveyed, limiting perspective on hierarchical differences. - HRM scope: Focused on human resource planning, training and development, and remuneration/benefits; omitted other HRM dimensions such as performance evaluation and non-financial rewards. - Cross-sectional design: Limits causal inference and temporal dynamics of knowledge sharing and innovation. - Respondent perception bias: All constructs measured via self-report; though reliability/validity checks were strong, common method variance cannot be entirely ruled out.