Communication strategies in the context of IoT-based smart farming are not uniform processes but consist of multiple interacting dimensions, including the credibility of information sources, the media through which information is delivered, the communication channels employed, and the degree of receiver participation (Prakash et al., 2023; Strong et al., 2022). These dimensions correspond to distinct communication forms commonly observed in agricultural innovation systems.

Face-to-face communication and offline training activities facilitate interactive learning and trust building, particularly during early stages of technology introduction (Cano et al., 2023). Farmer-to-farmer communication strengthens social learning and peer validation, while digital platforms and social media communities enable continuous information exchange and reinforcement. In contrast, formal policy communication primarily functions to legitimize innovation and reduce institutional uncertainty (Deperrois et al., 2023). This multidimensional perspective highlights that communication effectiveness depends on both the channel employed and the interaction intensity between communicators and farmers (Masambuka-Kanchewa et al., 2020; Spurk et al., 2023).

Communication plays a fundamental role in shaping how farmers interpret, evaluate, and eventually decide to adopt technological innovations. Studies demonstrate that communication effectiveness depends not only on information transfer but also on the credibility of sources, delivery channels, media formats, and capacity of receivers to interpret and operationalize technological knowledge (Kante et al., 2017). Farmer-to-farmer interactions, facilitated through digital platforms and social networks, are particularly influential, as farmers perceive peer-based information as more authentic and closely aligned with their context (Dilleen et al., 2023; Kusnandar et al., 2023). This aligns with the premise that innovation perception is socially constructed through shared communication experiences, not merely learned through formal training (Luqman and Van Belle, 2017; Yasmin and Akhter, 2023).

Research on IoT adoption highlights that communication that emphasizes economic gain, environmental benefits, and technological practicality reinforces positive perceptions of innovation attributes such as relative advantage, compatibility, simplicity, and convenience (Jayashankar et al., 2018; Vekariya et al., 2024). When supported by accessible communication platforms, these perception factors become strong predictors of adoption intention and behavior (Bahari et al., 2024; Yang et al., 2024). Conversely, ineffective communication increases uncertainty, particularly regarding data privacy, security risks, and technological dependence (Gupta et al., 2024; Wiseman et al., 2019). Therefore, innovation perception functions as the critical cognitive mechanism linking communication exposure to behavioral adoption.

Institutional support acts as a structural enabler in technology diffusion, particularly in resource-constrained farming environments. Support from agriculture offices, digital innovation management bodies, and farmer groups facilitates access to information, technology, training, and digital infrastructure (Sekele et al., 2025; Winarno et al., 2025). In contexts where IoT investment is financially limiting, institutional interventions such as subsidization, digital capacity building, and cluster-based farmer organization significantly reduce adoption barriers (Mariyono, 2019; Pasupuleti et al., 2025; Rayhan et al., 2024). Furthermore, sustained institutional engagement generates trust, helping mitigate farmers’ perceived uncertainty linked to technology complexity and cyber-related risks (McCaig et al., 2026; Russell et al., 2025).

Studies reveal that institutional support is most effective when embedded in participatory digital innovation ecosystems, where technology is not merely distributed but collectively interpreted and adapted to local practices (Bulut and Wu, 2024; Huo et al., 2024). These institutional mechanisms help shape innovation perception, indicating that adoption is not solely driven by technology availability but co-determined by how institutional actors frame and contextualize innovation within farmers’ operational realities (Peña-Holguín et al., 2025; Scur et al., 2023). Institutional support in this study encompasses the roles of agricultural bureaus, digital innovation authorities (Komdigi), and farmer organizations. These institutions provide support through multiple functional mechanisms, including policy facilitation and incentives, technical training and extension services, infrastructure provision, and organizational coordination at the local level.

Studies applying cognitive adoption models suggest that innovation perception functions as a mediating variable that translates external support structures into behavioral change (Arrang et al., 2025; Bahari et al., 2024). Farmers exposed to reliable institutional support and relevant communication are more likely to perceive IoT as operationally feasible and beneficial, which activates adoption intent and ultimately sustained use (Peña-Holguín et al., 2025). When perceptions regarding economic utility, compatibility with existing practices, technological clarity, and operational convenience are positively constructed, IoT adoption rates increase substantially (Dilleen et al., 2023; Yang et al., 2024).

Conversely, negative perceptions arising from communication gaps or institutional weaknesses—particularly related to technological reliability, financial viability, or data safety—remain strong inhibitors regardless of infrastructure availability (Narwane et al., 2022; Wiseman et al., 2019). This indicates that innovation perception acts as the cognitive filter through which structural support is evaluated, affirming its relevance as a mediator in the adoption process.

Existing studies predominantly examine communication and institutional support independently as drivers of innovation adoption. Yet, limited research empirically evaluates how these determinants interact simultaneously and how their effects are transmitted through innovation perception in the context of IoT-based smart farming (Bulut and Wu, 2024; McCaig et al., 2023). Further, the mediating role of innovation perception has been acknowledged conceptually yet remains underexplored empirically in agricultural digital transformation (Arrang et al., 2025; Peña-Holguín et al., 2025).

Thus, this study contributes theoretically by: Integrating innovation communication and institutional support into a unified structural model and empirically testing the mediating role of innovation perception in shaping farmer adoption behavior of IoT-intensive smart farming technologies. This research provides a behavioral and communication-centered perspective on digital agriculture, extending current socio-technical frameworks and offering strategic insights into how innovation adoption can be accelerated through communication and institutional alignment.

Institutional mechanisms—including agricultural offices, digital innovation management entities (Komdigi), and farmer group organizations—play a crucial role in shaping farmers’ access to information, training, technological infrastructure, and resources needed for innovation adoption. When institutions provide structured guidance and facilitative support, farmers tend to perceive digital innovation as more compatible with their farming practices, economically beneficial, and less risky (Sekele et al., 2025; Winarno et al., 2025). Institutional involvement further enhances innovation legitimacy by embedding it into local decision structures (Huo et al., 2024). Thus, stronger institutional support is expected to reinforce positive perceptions of IoT-based smart farming.

Effective communication—through credible sources, appropriate media, efficient channels, and receptive audiences—provides farmers with meaningful insights into technological innovation. When communication emphasizes relative profitability, adaptability, reduced complexity, and operational convenience, innovation perception tends to improve (Jayashankar et al., 2018). Digital peer communication further strengthens innovation understanding and reduces uncertainty (Dilleen et al., 2023). Therefore:

Institutional support lowers adoption barriers by facilitating access to digital infrastructure, training, and financial instruments (Pasupuleti et al., 2025; Rayhan et al., 2024). When farmers experience supportive engagement from institutional actors, they are more likely to adopt IoT technology and expand its application across farming activities (Bulut and Wu, 2024). Accordingly:

Communication enables farmers to observe demonstration results and practical use cases, shaping innovation intentions and behavioral outcomes (Luqman and Van Belle, 2017). Clear and persuasive innovation messaging reduces ambiguity related to duration and functionality of IoT usage (Yasmin and Akhter, 2023). Therefore:

Innovation perception—reflected in relative advantage, suitability, simplicity, and usability—is a recognized predictor of technology adoption (Arrang et al., 2025; Bahari et al., 2024). When perception is favorable, both the duration of technology use and number of features implemented are expected to increase.

Innovation perception acts as a cognitive filter connecting external enablers to behavioral action (Peña-Holguín et al., 2025). Institutional support and communication contribute to shaping innovation understanding, which in turn triggers adoption. Therefore, perception is expected to mediate the influence of communication and institutional mechanisms on adoption behavior.

The proposed research model (Figure 1) depicts the hypothesized relationships between constructs, aligning with the theoretical development above. Institutional support (X1) and innovation communication (X2) are posited as exogenous variables. Innovation perception (Y1) functions as a mediating construct, while innovation adoption (Y2) represents the endogenous outcome variable.

This study employed a quantitative survey as the primary method of analysis, supported by qualitative interviews conducted to provide contextual understanding of farmers’ experiences with IoT-based smart farming. The questionnaire was developed based on established literature and expert consultation, followed by refinement to ensure clarity and relevance to the study context. Qualitative interviews were used to contextualize survey findings and to enrich interpretation, rather than serving as a parallel dataset for triangulation. Qualitative data were collected through semi-structured interviews with agricultural extension agents, digital innovation facilitators, and local farming group leaders.

All constructs were operationalised as latent variables measured using reflective indicators based on validated scales. Institutional support was measured through items reflecting facilitation from agricultural offices, Komdigi, and farmer group organisations (Pasupuleti et al., 2025; Winarno et al., 2025). Communication variables assessed the credibility of information sources, effectiveness of media, channel relevance, and receiver engagement (Anwer Basha, 2025; Kante et al., 2017). Innovation perception captured relative advantage, compatibility, complexity, and convenience (Jayashankar et al., 2018; Vekariya et al., 2024; Yang et al., 2024), while innovation adoption was measured using indicators related to duration of use and the number of IoT features applied (Arrang et al., 2025). Responses were rated on a 5-point Likert scale (1 = strongly disagree; 5 = strongly agree).

The questionnaire was developed through a literature-based indicator synthesis and reviewed by three academic experts in agricultural digital innovation and communication. Items were refined for clarity and contextual relevance to West Java farming practices. A pilot test was conducted with 20 farmers of smart farming, resulting in minor linguistic adjustments to improve comprehension.

Data were collected in West Java Province, Indonesia, across 100 villages receiving IoT-based smart farming technology. Surveys were administered through face-to-face interviews to ensure response reliability given varied digital literacy levels. Qualitative interviews were held with four staff of West Java Provincial Agriculture Service, two staff of West Java Provincial Communication and Informatics Service, and 20 farmers to triangulate results.

The study population consisted of rice farmers actively involved in Gabungan Kelompok Tani (GAPOKTAN, farmer group associations), which function as the primary organizational units for agricultural coordination and technology dissemination at the local level. The population included both ordinary members and formal GAPOKTAN administrators—such as chairpersons, vice-chairpersons, secretaries, treasurers, and division heads—who are directly involved in decision-making, information dissemination, and technology adoption processes. The total population across the selected study areas comprised approximately 400 farmers.

To determine an appropriate sample size, Slovin’s formula was applied using a 5% margin of error, resulting in a minimum required sample of 200 respondents. This sample size is statistically adequate for Partial Least Squares Structural Equation Modeling (PLS-SEM) and satisfies established methodological recommendations for predictive models with multiple latent constructs, ensuring sufficient statistical power to detect medium effect sizes.

A purposive–stratified sampling strategy was employed to ensure balanced representation of farmers occupying different functional roles within GAPOKTAN. Stratification was based on organizational role (formal administrators versus general members) and involvement in smart farming communication and technology adoption activities. This approach ensured that respondents reflected both managerial perspectives—responsible for coordinating communication and institutional interaction—and operational perspectives, representing farmers as end-users of IoT-based smart farming technologies.

The study does not aim to estimate nationally representative adoption rates; rather, it seeks analytical generalization by examining relational mechanisms between communication strategies, institutional support, innovation perception, and adoption behavior within an institutionalized farmer group context. Given GAPOKTAN’s central role as a communication hub and institutional interface in rural Indonesia, this sampling design is appropriate for capturing the dynamics of communication-driven agricultural digital transformation.

Data were analysed using Partial Least Squares Structural Equation Modelling (PLS-SEM) in SmartPLS 4, given the model’s predictive orientation and mediating structure. A two-stage evaluation process was implemented: (1) measurement model assessment (indicator reliability, convergent and discriminant validity), and (2) structural model testing (path coefficients, f² effect sizes, R² adjusted, and Q² predictive relevance) (Hair et al., 2021). Bootstrapping with 10,000 subsamples was conducted to determine statistical significance using 95% bias-corrected confidence intervals.

Ethical procedures were strictly observed throughout the study. Ethical approval was granted by the Research Ethics Committee of Universitas Pendidikan Indonesia (Reference No. 109/UN40.K/PT.01.01/2025). All research activities adhered to internationally recognized ethical principles, including those outlined in the Belmont Report (1979) and the Declaration of Helsinki (2013 revision).

Participants were provided with detailed information sheets explaining the purpose, scope, risks, and benefits of the study. Written informed consent was obtained from all respondents, who were explicitly informed of their voluntary participation and their right to withdraw at any stage without penalty. Personal identities were protected through anonymisation using pseudocodes, and sensitive information was excluded from any verbatim quotations used in reporting.

A total of 200 respondents were selected from farmers participating in Gabungan Kelompok Tani (GAPOKTAN) across 100 Digital Villages implementing IoT-based smart farming initiatives. GAPOKTAN functions as a collaborative organizational structure consisting of farmer group leaders, committee members, and divisional coordinators, enabling coordinated decision-making and facilitating access to training, financial services, and innovation support.

The majority of respondents were male (70%), indicating a gender imbalance in technology decision-making, while women accounted for only 30% (Table 1). This disparity suggests that smart farming interventions may require gender-responsive strategies to enhance inclusivity. In terms of age distribution, 64% were between 39 and 53 years old, representing experienced but typically cautious technology adopters according to Rogers’ Diffusion of Innovations theory. Younger farmers (29–33 years) constituted only 2%, indicating limited youth engagement in agricultural digitization, while the oldest age group (59–63 years) represented 8%.

Regarding farming experience, 58% reported having 2–5 years, while 20% had more than 10 years. Farmers with shorter professional tenure may be more adaptable to innovation, whereas longer-established farmers are fewer but may function as influential opinion leaders. Most respondents (87%) were engaged in horticulture, where the need for precise input control supports strong alignment with IoT-based farming, whereas 13% operated in ornamental plant cultivation.

Land ownership data indicate that 86% cultivate plots smaller than 5,000 m², suggesting that IoT diffusion strategies must emphasize modular, low-cost solutions to support scalability among smallholder farmers. Collectively, the demographic profile highlights both opportunities and structural challenges in accelerating IoT adoption, particularly concerning age, digital literacy, gender inclusion, and farm scale.

Theoretically, this study advances existing knowledge by empirically verifying that communication functions not merely as an information-transmission instrument but as a cognitive-shaping mechanism that influences innovation perception and adoption behavior. The results extend technology adoption literature by demonstrating that perception is not solely shaped by economic or technical considerations but critically mediated by communication quality, structure, and trust mechanisms. Furthermore, the findings contribute to academic discourse by clarifying the conditional nature of institutional support, which becomes influential only when mediated through perceived relevance and value of innovation.

By integrating communicative and institutional determinants in a single model, this research introduces an interactional framework that moves beyond isolated analysis of policy or information dissemination. It responds to current scholarly calls for models that bridge socio-technical, communicative, and behavioral dimensions in smart farming adoption. The study therefore offers new empirical evidence supporting context-sensitive, perception-driven approaches in agricultural innovation research, particularly within digitally transitioning farming systems.

Practically, the findings suggest that adoption strategies for IoT-based smart farming should prioritize communication that is benefit-oriented, experience-based, and tailored to local decision-making dynamics. Communication efforts should leverage peer networks, farmer champions, and digital platforms to enhance trust and reduce perceived risks. Training and extension programs need to shift from purely technical instruction toward interpretative facilitation that helps farmers contextualize the relevance of IoT to their operations.

Institutional support should be structured not as a stand-alone driver but as a complementary mechanism that reinforces communication outcomes. Policy initiatives, subsidies, and infrastructure programs must be synchronized with field-level communication frameworks that translate institutional provisions into meaningful farmer insights. Financial incentives are likely to improve adoption outcomes only when supported by clear value-oriented communication and digital literacy strengthening. Additionally, capacity-building programs should incorporate ongoing technical assistance, participatory evaluation, and trust-building components related to data governance and cybersecurity concerns.

While this study provides important insights into the role of communication and institutional support in shaping innovation perception and adoption of IoT-based smart farming, several limitations should be acknowledged to contextualize the findings and inform future research.

First, Subsequent studies are encouraged to extend this model by incorporating channel-specific and content-specific indicators and applying multi-group analysis (MGA) to examine whether communication effects differ across distinct farmer groups. Comparative analyses may distinguish farmers with dominant exposure to face-to-face extension activities from those relying primarily on digital communication platforms, as well as groups predominantly receiving economic-benefit-oriented messages versus usability- or environment-focused communication.

Second, the study employed a cross-sectional design, which limits the ability to fully capture the temporal dynamics of perception formation and behavioral transition toward technology adoption. Since communication processes, institutional influence, and farmers’ innovation attitudes evolve over time, future research should consider longitudinal or panel-based approaches to explore how these relationships change across different stages of technology diffusion.

Third, the study is limited by its geographic focus and reliance on self-reported survey data. External factors such as market access, infrastructure availability, and climate risk were not explicitly modeled. Qualitative interviews were used for contextual insight rather than formal triangulation. Future studies should incorporate longitudinal designs, broader geographic coverage, and more diverse farmer profiles. Integration of qualitative and quantitative data through fully mixed-method approaches would further strengthen understanding of communication-driven technology adoption.