Traceability Model in an Agri-Food Chain: Application of Structural Equations

Abstract

Background: Given the disruption of the global market, technological factors are impacting food supply chains (FSCs). As a result, traceability is emerging as a potential solution for faster and more secure decision-making within FSCs. Methods: This research aims to define a traceability model for a shrimp food supply chain (FSC) in Ecuador using structural equation modelling which insists not only on the main factors that explain its overall performance but also on the effects of the changes in agents’ behaviours. The research was conducted between March and December 2025. A 41-item questionnaire was used for data collection and was administered to 73 stakeholders. The information was reduced and assembled in five main factors. A structural equation model was applied to explore the impact of agents’ coordination, digital transformations, and customer satisfaction on the traceability of the shrimp FSC. Results: The results show that customer satisfaction is broadly affected by the improvements in chain traceability. Furthermore, the results demonstrate the relevance of coordination, digitalization, and traceability as key factors for strengthening the FSC’s performance. Conclusions: The results could contribute to Sustainable Development Goals 12 and 17 and be applicable to other agri-food chains.

1. Introduction

The growing globalisation of food markets and the increasing complexity of supply chain networks have heightened the need for greater transparency, safety, and traceability [1]. Traceability is a multifunctional and multi-attribute concept that records activities occurring before, during, and after production, packaging, distribution, and delivery processes [2,3]. It therefore refers to the ability to obtain and retrieve detailed information on FSC activities, processes, and flows [2]. Traceability also helps minimise risk: when agreed standards are not met, it supports the identification and, where appropriate, the accountability of the responsible actors [4]. In this regard, Farina, Kocian [5] propose the concept of interoperable traceability applied to the FSCs.

The growing relevance of traceability in FSCs can be seen both in the majority segmented of mass production (‘food from nowhere’) and in increasingly differentiated foods based on often conflicting and sometimes competitive quality criteria (‘food from somewhere’) [6,7]. Specifically, in the mass production segment, Ecuador exports large quantities of shrimp produced in aquaculture facilities. The beginning of the Ecuadorian shrimp production dates back to the late 1960s, already having high levels of activity and significant global production by the late 1980s. In 2024, Ecuador was the world’s leading exporter of shrimp, exceeding 1.2 million tonnes, representing almost 32.5% of global shrimp exports [8]. Furthermore, in 2025 shrimp became Ecuador’s leading export product, surpassing oil for the first time. Ecuador exports to various markets, mainly Asian markets (China), but also to the United States and the European Union [9].

In agri-food productions as a whole, but especially in ‘food from nowhere’, the distance between production and consumption is particularly high. For this reason, traceability has become a business necessity, particularly given the growing demands for food quality and safety from intermediaries and final consumers [10]. The global landscape of the FSC is undergoing a significant transformation with the growing adoption of digital tools [11]. In the case of Ecuadorian shrimp, all major importers (China, the USA and the EU) have developed regulatory frameworks and industry agreements aimed at strengthening the shrimp traceability. In addition, the techniques which allow the reconstruction of the origin of production have been dramatically improved [12,13,14]. In its application of the ‘from farm to fork’ principle, the European Union has been particularly demanding [15]. This collection of changes in major consumer countries has led to a myriad of adjustments in the management systems of many leading producers, such as Bangladesh, Indonesia, and Vietnam [16,17,18].

Certification is another important component of food supply chain governance. Third-party certification schemes can serve as market and governance instruments by linking sustainability commitments with verifiable product information. A systematic review by Hilmi et al. [19] shows that certification can support socioecological responsibility and create opportunities for price premiums. However, the authors also emphasise that consumer recognition, credibility, and full traceability remain decisive factors. Although certification schemes vary across commodities, this insight is relevant to shrimp FSCs because claims related to origin, safety, and sustainability must be supported by reliable information throughout the chain.

Information technology reinforces this function by improving the availability, exchange, and use of data among FSC actors. Nugroho [20] notes that agriculture increasingly incorporates the Internet of Things (IoT), sensors, robotics, AI, intelligent supply chains, efficiency, and supply chain management. At the macro level, the same study finds that information globalisation increases agricultural producer prices in both developing and developed countries. Digital tools should therefore not be treated as an end in themselves, but as enabling mechanisms whose contribution depends on their adoption, integration, and consistent use across the FSC.

The reviewed academic literature approaches traceability from different perspectives. Sometimes, traceability is analysed focusing on aspects related with food safety and quality [21]. In this context, [22], have developed a food management tool for traceability in seafood FSCs. The tool’s guidelines focus on traceability and blockchain technology (BCT) which provides a decentralised structure for data collection guaranteeing their immutability. Also, BCT is a particularly powerful tool for guaranteeing payment security and transparency regarding product origin, certifications and contamination risks [23]. In this sense, as the potential effects of BCT on traceability are particularly relevant, BCT has possibly become the most studied tool in terms of its effects on FSC traceability [24].

Other studies have focused on different features. Thus, some analyses have developed a systematic technique for evaluating information losses, modifying and integrating modal analysis of failures, effects and criticality [25]. The study uses data derived from Bangladesh shrimp FSCs, comprising a farm, a warehouse, and a processor. By providing standardised visualisation and quantitative assessment of information losses, the proposed method offers a systematic approach for the analysis of vulnerabilities in the traceability of the shrimp FSC.

Another study analyses the perceived facilitators and barriers influencing the behavioural intentions of multiple stakeholders in the fruit FSC to adopt blockchain-based traceability technologies [26]. In this regard, BCTs are analysed considering obstacles, barriers and challenges associated with their adoption as well as the overall changes promoted in the governance of the chain [27]. In this context, others offer a comprehensive overview of digital transformation trends in food safety and their influence on traceability [23].

However, limited research has examined the interrelationship between technological elements, coordination and customer satisfaction from the perspective of actor and FSC traceability. This gap is directly addressed in the research presented in this article and, at the same time, becomes the theoretical contribution of the study. Our emphasis is therefore on analysing how explicit and tacit codes of coordination among agents not only affect the effective levels of traceability in the shrimp FSC in Ecuador, but also the effects that the development of traceability protocols have on the behaviour of shrimp FSC agents and, above all, on the satisfaction of final consumers. In other words, this article analyses the recursive and systemic effects that traceability can have on FSC governance (through changes in the behaviour of agents) and on consumer purchasing behaviour (based on an analysis of their satisfaction). All of this must also be placed in a broader context in which the improvement in shrimp FSC traceability must be understood in the context of rising concerns about identity, safety and sustainability of production facilities and processes [28].

On the other hand, despite the existence of some studies on shrimp FSCs, few have been conducted in Ecuador. This is particularly surprising for several reasons. Firstly, as already mentioned, Ecuador is currently the world’s leading exporter of shrimp. Secondly, the regulatory framework for shrimp production has changed in Ecuador, although possibly less so than in other producing countries. Therefore, there is still a wide margin for action and improvement [29].

These factors are added to the disruption of the global market in shrimp supply chains, caused by geopolitical conflicts [30] (wars and the collapse of logistical infrastructure that affect global exports and increase prices [31]), climate risks and diseases [32] (temperature differences due to climate change and the increase in shrimp diseases), and supply chain structure [33] (the lack of visibility and communication among supply chain actors, along with dependence on global markets). The academic literature acknowledges the existence of growing risks in the operations of agri-food companies and proposes a set of criteria designed both to minimise and to address these risks [34]. In this sense, shrimp supply chains face significant disruptions due to different factors; therefore, the study of these chains, such as the Ecuadorian one, is highly relevant for defining future adaptability strategies in these contexts. At the same time, these elements increase the vulnerability of supply chains to crises, and consequently negatively impact the Ecuadorian shrimp supply chain.

In this article, a traceability model is defined through the application of a systemic approach that takes the form of a system of structural equations which can be considered a significant contribution to the comprehensive analysis of the operation of the FSC shrimp industry in Ecuador, a contribution of this research.

2. Traceability from Theory and Hypothesis Development

Based on this background information, the following scientific propositions have been reached:

A. Coordination among actors in the FSC influences both the traceability of the chain as a whole and that of each individual actor.

Coordination within FSCs is a particularly challenging issue. Coordination among customers and suppliers is complex because it involves a broad set of interrelated flows and processes [35,36]. The forms of coordination inside the FSCs depend on various factors, including the dominant management models specifically in those organisations with the greatest drivenness [37]. In this sense, models of management that pursue very high levels of efficiency—by reducing errors and wastes as well as fostering continuous improvements—can have a systematic impact on increasing efficiency in FSCs [38].

In this regard, the introduction of traceability in an FSC involves a change in perspective. In this sense, it means understanding the effects that changes in specific organisations have on the collective behaviour of FSCs [38]. But it also involves understanding how changes in coordination mechanisms affect the operating models of specific companies. In this way, the analysis of traceability helps to reorient the focus from the internal processes and flows of specific agents to the combination of the processes and flows of multiple agents. Also, these processes must be aligned with the chain’s overall operation in order to satisfy final customers [35,39]. Those transformations imply greater information exchange among FSC agents. For example, coordination within the FSC would help reduce the bullwhip effect, high inventory costs and supply uncertainty [40]. This increase in available information and coordination among actors also results in a significant improvement in risk management and greater resilience in food supply chains [34,41].

Additionally, one of the limitations to achieving traceability in a chain is the security and availability about jointly generated data [27,42,43]. This demonstrates the need to strengthen collaboration among agents to contribute to the traceability of the chain [40,44,45,46].

The accuracy and fluidity of information flows between FSC participants are among the objectives of traceability [5,47]. Given that different actors use their own coding protocols and technologies, the main obstacles to achieving these objectives are associated with the variety of data coding and transmission methods. For this reason, [5] propose the concept of interoperability which must be understood as the ability of systems to share and use information jointly. This concept has several dimensions, for instance organisational, legal, semantic and technical [48]. FSCs characterised by strong coordination have been shown to be more likely to have interoperable traceability systems than those with weak coordination [5].

Based on these elements, the following hypotheses are defined:

H1. 

Coordination among agents directly influences the traceability in the overall FSC.

H2. 

Coordination among agents directly influences the traceability of the different agents who operate inside the FSC.

B. Digital technology promotes traceability in the agri-food chain

Digital transformations can influence FSC operations [3,23,49]. This is largely due to the need to manage flows and processes in real time to improve supply chain performance [50,51]. This is associated with the necessity of identification of the product’s upstream and downstream routes in order to achieve traceability and provide information to customers.

In this context, BCT is a solution that promotes traceability [3,52,53,54]. BCT is defined as an innovation that generates new information and communication technologies by supporting highly available, low-cost, distributed, decentralised and transparent data management that is resistant to manipulation [23,55]. The BCT becomes a traceability solution that can contribute to make more sustainable, resilient and efficient FSCs [3]. Therefore, the BCT comprises a set of tools able to generate long-term, searchable, and immutable public record archives [3]. In addition, the BCT provides a comprehensive framework for FSC agents who wish to achieve traceability [55]. Thus, BCT contributes to achieving integration into the FSC and traceability [50].

Another way of promotion of traceability is through the application of artificial intelligence (AI) [3,52,56]. Thus, some studies have explored how the use of AI promotes the advancement of Food Industry 4.0 [42]. At the same time, the use of AI not only is closely related with but also drives automation, real-time monitoring and the interconnection of supply chains [42].

In this sense, Halder, Islam [43] develop a structured taxonomy of AI-driven security mechanisms. Thus, they demonstrated how the use of AI contributes to improving traceability in the FSC, as well as identifying its limitations. Rossi, Gemma [1], through analysis traceability in wine, garlic and coffee FSCs, provide a roadmap for food traceability systems, covering everything from legal requirements to technological and analytical perspectives. Their analysis deals with topics related with innovative analytical technologies, as well as emerging digital technologies such as AI.

In this regard, a baseline study for this research is that of Khan [3] which analyses how BCT, the social Internet of Things, and AI affect supply chain traceability. Although the variable of the Internet of Things was not used in our research because of the low level of technological development of the Ecuadorian shrimp FSC, an initial exploratory approach was made in the case of AI. Although there are preliminary and laboratory-level studies of the Internet of Things at some points in the chain in Ecuador, these potential innovations are not yet being applied to actual production processes [57]. In any case, part of our study focuses on the effect of digital transformations, such as blockchain technology and artificial intelligence techniques in Ecuadorian shrimp FSCs, on achieving traceability. The following hypothesis is, therefore, proposed:

H3. 

Digital transformations influence traceability in the FSCs.

C. Traceability in FSCs affects customer satisfaction.

One of the main objectives of actors interacting within an FSC is to satisfy customers [21]. This objective can be assessed through different performance indicators. Within specific socio-technical systems, FSC traceability can be associated with innovations that improve operational performance and, consequently, customer satisfaction [40].

Another view of this relationship is that established by Khan [3] who considers that the design of the chain affects the degree of visibility of orders. This latter element is an important aspect of traceability. He also mentions the need for customers to track the history and origin of products in order to decide whether to purchase them, thereby guaranteeing product integrity and safety. These concerns coincide with those of many other authors [3,51,53]. Furthermore, the importance of information, visibility, and transparency for achieving traceability and improving customer satisfaction is also often highlighted [49].

Similarly, ref. [58] explore, based on consumer knowledge, preferred information and willingness to pay for the traceability of processed foods. As a result, they demonstrated that customers, even if they are unfamiliar with the term, associate traceability with product safety and quality. This demonstrates that it is necessary to increase customer consciousness in order to ensure that their purchasing decisions are based on FSC traceability. Thus, the consumption of products that strengthen and support such traceability could be promoted [53,54,58], mentions how the naturalness of the product is associated with FSC traceability. In this sense, closeness to nature is associated with producers who, if not local, are at least identifiable. This latter aspect reinforces ecological trends by increasing customers’ guarantees.

In this context, ref. [54] examine the impact of consumer knowledge on the adoption of BCT in a supply chain comprising a manufacturer and a retailer. They develop different scenarios considering as criteria the different levels of consumer knowledge about traceability and focusing the analysis on those consumers who are aware of the advantages of these technologies. The results reveal that the adoption of BCT improves the overall performance of the supply chain and makes it more sensitive to the level of consumer knowledge about traceability.

Based on the above elements, the following scientific proposition is defined.

H4. 

Traceability in FSCs influences customer satisfaction.

Figure 1 presents the conceptual framework of this study, derived from the previous theoretical discussion.

3. Materials and Methods

3.1. Classification of the Research

This study analyzes traceability in the Ecuadorian shrimp FSC using structural equation modelling. Its cross-sectional design involves observing FSC actors at a single point in time, allowing an assessment of traceability practices during the study period while acknowledging that changes over time are beyond the scope of the analysis [59].

3.2. Data Collection and Initial Processing

The shrimp chain consists of 200 actors according to a preliminary study developed in the framework of the research collaborative project entitled ‘Circularity in the export shrimp chain in Manabí, Ecuador’ (PYTAUTO2966-2023-IPG0004) (

Table 1. Phases and stages of the Ecuadorian shrimp food chain.

Phase	Step	Quantity
Primary production	Internal larval rearing	29
	Export larval rearing	9
	Production and cultivation	93
Storage	Collection centre	4
Secondary production	Shrimp production: Heading deveining, peeling	22
	Production of value-added products from shrimp waste	2
Sale	Processing for the local market	14
	Direct and indirect exports	27
Total		200

Source: Own elaboration.

). This therefore is the size of the population of the Ecuadorian shrimp FSC. This population is divided into different types of organisations (Table 1).

This research was conducted using a quantitative approach, based on primary data collected through an online questionnaire. The instrument was designed based on a systematic review of the specialised literature on traceability and agri-food supply chains and was specifically adapted to the objectives and characteristics of this study. As a result of this process, minor adjustments were made to the wording of the questionnaire to ensure its proper understanding within the context analysed. In addition, a pilot survey was conducted to verify the suitability of the questions from the respondents’ perspective. The target population of the survey consisted of individuals in management positions (owners, senior managers, middle managers, and operations managers) in companies that are part of the Ecuadorian shrimp FSC, covering its linkages from primary production to marketing and export.

Data collection took place between March 2025 and December 2025. A non-probabilistic convenience sampling method was used. As is usual in this type of research, the selection made (in this case, by convenience) is assumed to have patterns similar to those made according to purely random criteria. Thus, a size of 51 individuals is required for the sample to have a margin of error of 10%. However, a total of 73 questionnaires was obtained that were considered complete and valid. Considering the total population of 200 organisations, this is a fairly high response rate of 36.5%. The distribution of questionnaires by type of organisation based on company size is shown in

Table 2. Distribution of the sample by company size.

Size of the Company	Frequency	Percent
Large enterprise (200 or more employees)	13	17.8
Medium-sized enterprise A (50 to 99 employees)	20	27.4
Medium-sized enterprise B (100 to 199 employees)	15	20.5
Micro enterprise (1 to 9 employees)	19	26
Small enterprise (10 to 49 employees)	6	8.2
Total	73	100

Source: Own elaboration.

.

The fieldwork was affected by security conditions in Ecuador during the data-collection period. Some managers and organisational representatives were reluctant to participate in surveys or share internal operational information because of confidentiality, security, and organisational-exposure concerns. Despite repeated contact efforts, access to respondents was therefore limited.

3.3. Measuring Instruments

Traceability in the FSC was assessed using five factors and 41 questionnaire items that correspond to the conceptual framework presented in Figure 1 and follow the theoretical guidelines of [3,40]. Appendix A provides a detailed description of the variables, the associated questions, and their theoretical foundations. The five constructs are described below:

Coordination (COO): Coordination with buyers and suppliers [40,46]. This factor encompasses 12 questions from the questionnaire.

Supply chain traceability (SCT) based on Khan [3], Bosona and Gebresenbet [45], Khan, Lee [60], Khan, Parvaiz [61]: This construct includes 9 questionnaire items.

Actor-level traceability (ALT) defined by Shou, Zhao [40]: This factor encompasses 4 questions from the questionnaire.

Customer satisfaction (CS), based on Shou, Zhao [40], Bozarth, Warsing [62]: This construct includes 6 questionnaire items.

Digital transformations (DTs), including AI [3,52,63] and BCT elements [3,52]: This factor encompasses 10 questions from the questionnaire.

The constructs were derived from the academic literature. The empirical adequacy of their indicators was subsequently assessed through factor analysis and measurement-model evaluation, as explained in Section 4. The instrument uses a Likert scale from 1 to 5, where Very Low = 1, Low = 2, Medium = 3, High = 4, Very High = 5.

Factor analysis was used as a data-reduction technique with the aim of identifying the underlying dimensions that explain the structure of the variables analysed in the traceability study of the supply chain under investigation. For the factor analysis estimation, the Principal Component Analysis (PCA) technique was employed, which is widely used in exploratory studies when seeking to reduce a large set of variables to a smaller number of representative factors.

Subsequently, in order to facilitate the interpretation of the factors obtained, a Varimax rotation was applied. This procedure maximises the variance of the factor weights within each factor, resulting in a clearer structure and making it easier to identify the groups of variables that constitute each dimension. As a result of the analysis, the 41 questions in the questionnaire were grouped into five main factors, consistent with the study’s conceptual framework and supported by the previous academic literature.

3.4. Data Analysis

SmartPLS Version 4.1.1.6 software was used to process and analyse the information, applying the partial least squares structural equation modelling (PLS-SEM) technique. This methodological approach was chosen due to its analytical robustness and its widespread use in studies on FSCs. Thus, structural models allow examining recursive and systemic effects without requiring strict assumptions of normality in the data [64] nor the availability of large samples [65].

This study employed structural equation modelling (SEM), based on the use of previously defined and calculated factors. SEM is a multivariate statistical approach widely used in the social sciences and in management and supply chain studies to analyse complex relationships between latent and observed variables. SEM allows for the simultaneous evaluation of measurement models and structural models, enabling the analysis of both the validity of theoretical constructs and the causal relationships between them [64].

Within the SEM approach, there are two main methodological strands: covariance-based SEM (CB-SEM) and partial least squares SEM (PLS-SEM). Whilst CB-SEM is primarily used to confirm well-established theories and requires strict statistical assumptions, such as multivariate normality and relatively large samples, PLS-SEM is geared towards predictive and exploratory analysis, being particularly suitable when the aim is to develop theories, analyse complex models or work with moderate-sized samples [64].

The methodological choice of PLS-SEM was precisely motivated by its suitability for exploratory and predictive research with moderate sample sizes and complex models. According to [64,65], PLS-SEM is appropriate for studies with relatively small samples, particularly when the research objective is theory development and prediction rather than strict covariance confirmation. Accordingly, the use of PLS-SEM in this study is justified by the exploratory and predictive nature of the research, the complexity of the proposed structural model, and the moderate sample size obtained from the Ecuadorian shrimp food supply chain.

4. Results

A descriptive analysis was conducted for the variables grouped into the five factors described in Section 3.3.

Table 3. Descriptive analysis.

Factor	Item	Mean	Standard Deviation
Coordination (COO) Mean: 3.041	COO01	3.233	1.309
	COO02	3.082	1.322
	COO03	3	1.271
	COO04	2.89	1.223
	COO05	3.082	1.352
	COO06	2.712	1.277
	COO07	3.301	1.143
	COO08	3.192	1.094
	COO09	3.096	1.112
	COO10	3.041	1.039
	COO11	3.014	1.176
	COO12	2.849	1.119
Supply Chain Traceability (SCT) Mean: 3.359	SCT01	3.274	1.162
	SCT02	3.301	1.257
	SCT03	3.616	1.278
	SCT04	3.438	1.26
	SCT05	3.493	1.273
	SCT06	3.219	1.337
	SCT07	3.178	1.358
	SCT08	3.411	1.28
	SCT09	3.301	1.289
Actor-level Traceability (ALT) Mean: 3.212	ALT01	3	1.228
	ALT02	3.137	1.317
	ALT03	3.438	1.182
	ALT04	3.274	1.219
Customer Satisfaction (CS) Mean: 3.810	CS01	3.836	1.194
	CS02	3.767	1.222
	CS03	3.973	1.227
	CS04	3.753	1.168
	CS05	3.753	1.211
	CS06	3.781	1.184
Digital Transformations (DTs) Mean: 2.743	DT01	2.959	1.308
	DT02	3.055	1.333
	DT03	2.918	1.431
	DT04	3	1.385
	DT05	2.932	1.388
	DT06	2.301	1.213
	DT07	2.411	1.312
	DT08	2.808	1.341
	DT09	2.575	1.097
	DT10	2.466	1.261

Source: Own elaboration.

reports the mean and standard deviation for each item.

Based on the average values observed, customer satisfaction is identified as the factor most highly valued by the participating agents, with averages close to and above 3.75 on a scale of 1 to 5. This suggests that shrimp FSC agents perceive favourable performance in terms of meeting expectations, product quality and customer response.

For FSC traceability, item means range from 3.18 to 3.62, indicating a moderate level of implementation across the different stages of the Ecuadorian shrimp FSC. Although product-tracking and tracing mechanisms are present, their application is not fully consistent or standardised.

Actor-level traceability shows similar values, with means ranging from 3.00 to 3.44. These results suggest that traceability efforts are concentrated primarily within individual organisations rather than fully integrated at the inter-organisational level.

In this regard, coordination among FSC agents shows average values close to 3.0, with some variation between items. This suggests the existence of basic coordination mechanisms, although these are still limited in terms of strategic alignment, systematic information exchange, and synchronisation of productive processes.

Finally, digital transformations received the lowest overall ratings, with item means ranging from 2.30 to 3.05. This result indicates a limited level of adoption of advanced digital technologies, including blockchain and AI, and points to a technological gap within the FSC.

Measurement Model Quality

Before assessing the measurement and structural models, the adequacy of the data for dimensionality reduction was evaluated. The Kaiser–Meyer–Olkin measure reached a value of 0.904, indicating excellent sampling adequacy, while Bartlett’s Test of Sphericity was statistically significant (χ² = 4286.368; df = 820; p < 0.001). These results confirm that the correlation matrix was suitable for factor analysis.

Principal Component Analysis (PCA) with Varimax rotation was subsequently applied to the 41 original indicators. Although four components exceeded the eigenvalue-greater-than-one criterion, a five-factor solution was retained because the constructs had been theoretically specified a priori in the conceptual framework. This decision was also supported by the high cumulative variance explained by the five-factor solution, which reached 83.59% of the total variance. The retained structure corresponds to coordination, supply chain traceability, actor-level traceability, customer satisfaction, and digital transformations.

This procedure allowed the study to verify that the empirical structure of the indicators was consistent with the theoretically proposed dimensions, while recognising that the retention of the fifth factor was theory-driven rather than based exclusively on the eigenvalue criterion.

The measurement model was initially assessed through the examination of the external loadings of the reflective indicators. During this process, the indicator DT04, belonging to the digital transformations construct, presented an external loading of 0.547, which was below the recommended threshold of 0.70. Consequently, this indicator was removed from the final measurement model in order to improve indicator reliability and convergent validity. The final model was therefore estimated using 40 retained indicators.

After item purification, all retained indicators exhibited external loadings above the recommended threshold, ranging from 0.795 to 0.972 [54]. Internal consistency reliability was assessed using Cronbach’s Alpha and Composite Reliability. All constructs exceeded the recommended threshold of 0.70, with Cronbach’s Alpha values ranging from 0.937 to 0.979 and Composite Reliability values ranging from 0.955 to 0.983 (

Table 4. Measurement model reliability and validity.

Factor	Item	Loading/Weights	t-Value	CA	CR	AVE
Coordination (COO)	COO01	0.866	30.012	0.968	0.969	0.725
	COO02	0.875	33.092
	COO03	0.885	39.920
	COO04	0.887	37.362
	COO05	0.889	39.338
	COO06	0.795	15.386
	COO07	0.820	16.487
	COO08	0.869	23.670
	COO09	0.823	14.608
	COO10	0.844	20.753
	COO11	0.840	17.268
	COO12	0.813	14.607
Supply Chain Traceability (SCT)	SCT01	0.905	37.087	0.968	0.972	0.792
	SCT02	0.937	59.276
	SCT03	0.811	14.242
	SCT04	0.912	42.761
	SCT05	0.904	39.164
	SCT06	0.864	28.030
	SCT07	0.899	27.547
	SCT08	0.856	23.495
	SCT09	0.915	40.406
Actor-level Traceability (ALT)	ALT01	0.932	51.233	0.938	0.955	0.842
	ALT02	0.940	76.098
	ALT03	0.880	27.923
	ALT04	0.918	43.644
Customer Satisfaction (CS)	CS01	0.953	63.421	0.981	0.983	0.907
	CS02	0.940	59.091
	CS03	0.946	60.428
	CS04	0.945	60.842
	CS05	0.960	92.958
	CS06	0.972	71.162
Digital Transformations (DTs)	DT01	0.839	64.256	0.976	0.971	0.760
	DT02	0.818	84.092
	DT03	0.895	91.053
	DT05	0.898	81.996
	DT06	0.931	23.102
	DT07	0.939	15.366
	DT08	0.951	40.713
	DT09	0.960	4.944
	DT10	0.945	30.922

Note: All loadings are significant at p < 0.001 (two-tailed, based on bootstrapping results).

).

Second, internal-consistency reliability was assessed using Cronbach’s Alpha (CA) and Composite Reliability (CR). As shown in Table 4, CA values exceeded the recommended threshold of 0.70, indicating high levels of internal consistency [66]. CR values were also above 0.70, which is considered adequate in PLS-based studies [64].

Third, convergent validity was analysed using the average variance extracted (AVE). According to Hair, Sarstedt [64], AVE values should be equal to or greater than 0.50 for the construct to be considered to adequately explain the variance of its indicators. In this study, AVE values ranged from 0.725 to 0.907, confirming a high level of convergent validity in the constructs analysed. Although excessively high values (≥0.95) may indicate redundancy between items, none of the constructs exceed this threshold. Consequently, no problems of redundancy or loss of conceptual validity are identified (

Table 5. Discriminant validity of the measurement model (HTMT ratio).

Relationship Between Constructs	HTMT	Mean	2.5%	97.5%
Coordination ↔ Actor-level traceability	0.929	0.929	0.870	0.974
Customer satisfaction ↔ Actor-level traceability	0.692	0.688	0.496	0.840
Customer satisfaction ↔ Coordination	0.669	0.663	0.468	0.812
Digital transformations ↔ Actor-level traceability	0.736	0.736	0.562	0.869
Digital transformations ↔ Coordination	0.828	0.828	0.718	0.908
Digital transformations ↔ Customer satisfaction	0.435	0.431	0.230	0.612
Supply chain traceability ↔ Actor-level traceability	0.796	0.792	0.654	0.900
Supply chain traceability ↔ Coordination	0.799	0.796	0.658	0.895
Supply chain traceability ↔ Customer satisfaction	0.840	0.837	0.739	0.910
Supply chain traceability ↔ Digital transformations	0.733	0.733	0.609	0.832

Source: Own elaboration.

).

Fourth, discriminant validity was assessed using the heterotrait–monotrait ratio (HTMT) criterion in Table 5. Following [67], an HTMT threshold below 0.90 was considered, together with verification that the 95% confidence intervals obtained through bootstrapping did not include the value 1. The results show that most of the relationships between constructs have HTMT values clearly below the recommended threshold. Although the relationship between coordination and traceability at the actor level reaches an HTMT value slightly above 0.90, its confidence interval (0.870–0.974) does not include the unit, which allows us to statistically confirm the discriminant validity under the interval approach.

Regarding the overall fit of the model, the indicators show adequate performance. The Goodness of Fit (GoF) value is 0.53, exceeding the threshold of 0.36, which indicates a high level of fit for the proposed model [68]. Likewise, the SRMR index has a value of 0.081, which is below the recommended limit of 0.10, suggesting an adequate correspondence between the observed and estimated correlation matrices [67]. In addition, the explained variance values (R²) exceed the recommended minimum of 0.10, highlighting chain traceability, actor traceability and customer satisfaction, which confirms the explanatory power of the model in the context of the shrimp chain.

Once these criteria had been verified, a bootstrapping procedure was applied with 5000 resamples to test the hypotheses (

Table 6. Hypothesis testing.

Relationship		Path Coefficient	t-Value	p-Value	Supported
H1	COO → SCT	0.562	4.392	<0.001	Yes
H2	COO → ALT	0.890	35.691	<0.001	Yes
H3	DT → SCT	0.244	1.878	0.060	Not Supported
H4	SCT → CS	0.818	18.438	<0.001	Yes

Source: Own elaboration. Note: The hypothesis tests were estimated through a bootstrapping procedure with 5000 resamples. The final model was estimated after removing the indicator DT04 due to its insufficient external loading.

). Coordination among FSC actors has a positive and significant effect on both FSC traceability and actor-level traceability, confirming the importance of inter-organisational coordination in the shrimp export FSC. The relationship between digital transformations and FSC traceability is positive but not statistically significant at the 5% level (β = 0.244; t = 1.878; p = 0.060). H3 is therefore not supported.

Finally, the results show that FSC traceability has a positive and significant impact on customer satisfaction, highlighting that greater levels of visibility, transparency and information directly contribute to the improvement of the product perception of final customers. Taken together, these findings confirm the empirical validity of the proposed structural model and support the relevance of coordination and traceability as key factors in strengthening the performance of the shrimp FSC.

5. Discussion

The findings show that coordination and traceability are central elements in the performance of the Ecuadorian shrimp FSC. The confirmation of H1 and H2 indicates that coordination among actors significantly influences both supply chain traceability (β = 0.562; p < 0.001) and actor-level traceability (β = 0.890; p < 0.001). These results are consistent with [46,67], who argue that process integration and information exchange form the basis of effective supply chain management. They also align with [40], who show that the alignment of coordination and traceability promotes operational innovation and performance, and with [5], who argue that stronger coordination increases the likelihood of developing interoperable traceability systems. The results therefore suggest that organisational alignment is not merely an operational capability, but a structural condition for systemic traceability.

In contrast, the relationship between digital transformations (DTs) and supply chain traceability (SCT) is positive but not statistically significant at the 5% level (β = 0.244; t = 1.878; p = 0.060; 95% CI [−0.011, 0.487]). Therefore, H3 is not supported in the final model. This finding does not imply that digital technologies are irrelevant. Rather, the evidence from this sample is insufficient to confirm a significant direct effect. The result differs from frameworks that emphasise the potential of BCT, the social Internet of Things, and AI to improve transparency and traceability [3,52]. In the Ecuadorian shrimp FSC, the descriptive results suggest that the limited adoption and integration of advanced digital tools may help explain why the direct effect is not yet statistically significant.

Finally, the confirmation of H4 (β = 0.818; p < 0.001) demonstrates that traceability has a major significant impact on customer satisfaction. This assertion is in line with [21], who link traceability with perceived safety and quality, and is coherent with [54], who point out that consumer knowledge about traceability increases the perceived value of the product. Furthermore, while studies like [69] show that traceability primarily drives sustainability performance through green logistics, our results confirm that its benefits extend directly to market perception and consumer trust. When comparing our work comprehensively with previous structural modelling approaches, distinct gaps emerge. Research such as that by [70] applied structural equation modelling to analyse traceability mainly in relation to isolated organisational performance or sustainability outcomes. Similarly, refs. [71,72] highlight the importance of supply chain visibility and flexibility for business performance. In our model, these concepts are integrated and expanded from a broader, simultaneous perspective.

Overall, the main contribution of this study lies in empirically demonstrating that coordination and traceability are central elements in explaining customer satisfaction within an FSC. Although digital transformation was incorporated into the systemic framework, its direct effect on supply chain traceability was not statistically significant in the present sample. Digital tools should therefore be understood as potential enabling mechanisms whose contribution may depend on their level of adoption, integration, and consistent use across the chain. By modelling these relationships simultaneously using PLS-SEM, this research provides valuable empirical evidence in a scarcely explored developing-country context and opens avenues for future research on the integration of sustainability indicators, supply chain visibility, and the IoT.

6. Conclusions

This research explores and examines in depth the nature of the processes that explain the relevance of traceability as an emerging criterion in FSC governance. This research presents a case study based on a structural equation model that evaluates the relationships among agents’ traceability, coordination, AI, BCT, and customer satisfaction in Ecuador’s shrimp export supply chain. A questionnaire defined in previous research (Appendix A) was administered to 73 FSC stakeholders. A set of hypotheses based on the questionnaire variables and theoretical frameworks was then defined and tested using SEM analysis.

The results show that coordination among actors positively and significantly influences both supply chain traceability (H1) and actor-level traceability (H2). Supply chain traceability also has a positive and significant effect on customer satisfaction (H4). By contrast, the relationship between digital transformations and supply chain traceability is positive but not statistically significant at the 5% level; H3 is therefore not supported. Digital tools should consequently be understood as potential enabling mechanisms rather than as a statistically confirmed direct driver of traceability in the present sample. Their contribution may depend on the level of adoption, integration, and consistent use across the chain. These findings highlight the importance of strengthening coordination and traceability practices among FSC stakeholders and of developing public policies that support food safety, information exchange, and supply chain governance.

Following validation of the model, it is concluded that coordination among supply chain actors is fundamental for both chain traceability and customer satisfaction. This aligns with the fundamental objective of a supply chain—customer satisfaction—and with the need for integration among the different actors [73] to adapt to disruptive changes in the global market [6]. This constitutes a challenge for the supply chain under study. Based on these conclusions, improvements in the practices of FSC stakeholders are recommended to enhance traceability and coordination. Furthermore, there is a need for public policies that promote the development of these practices, such as a food safety and supply chain law.

Despite its structural contributions, this study possesses boundaries that limit the immediate generalizability of the findings. First, the cross-sectional design captures a specific snapshot in time, and the non-probabilistic convenience sampling—while practically necessitated by severe public safety and strict corporate confidentiality crises in Ecuador during 2025—means results should be translated to other regions with caution. Second, although the sample size (n = 73) is statistically robust for PLS-SEM estimations, it represents a specific segment of the regional shrimp industry. Third, the data relies on self-reported measurements from management, which may introduce subjective perceptions regarding traceability performance. To address these limitations and expand the scientific scope, future research should go beyond traditional frameworks. While it remains highly relevant for future works to directly include operational variables such as performance, visibility, and IoT in the models, as well as explicitly testing the direct relationship between the actor-level traceability variable and the customer satisfaction variable, new strategic opportunities must also be addressed. Future agendas should lie in establishing longitudinal frameworks to track traceability evolution over time, integrating objective digital telemetry and blockchain transaction logs to effectively mitigate self-reporting bias, and developing cross-commodity comparative studies within the wider Latin American agri-food ecosystem to enhance overall model generalizability.