Data-Driven Performance Evaluation and Behavior Alignment in Port Operations: A Multivariate Analysis of Strategic Indicators

Abstract

This study investigates how a KPI-based incentive system can enhance behavior alignment and operational performance in complex port environments, addressing the growing need for data-driven governance in public logistics operations. The research analyzes a two-year dataset from 145 navigation teams across 22 Brazilian ports, employing a multivariate approach that integrates cluster analysis, Wilcoxon signed-rank testing, and exploratory factor analysis (EFA). Results reveal distinct performance tiers, with high-performing ports showing consistent improvement in safety, compliance, and technical availability, while low-performing areas exhibit stagnation or decline. The stability of the three-factor structure—Operational Safety, Technical Availability, and Procedural Compliance—over time confirms the model’s robustness and suggests increasing internalization of performance expectations by teams. These findings demonstrate that incentive systems, when combined with transparent feedback and recognition, can align team behaviors with organizational goals and drive sustained operational improvement. The study contributes a validated, replicable framework for performance management, offering practical guidance for enhancing governance and strategic alignment in high-risk, complex settings. Future studies should explore motivational factors and the model’s applicability across the public sector.

1. Introduction

Modern organizations face demanding situations to improve their competitiveness and sustainability, and increase their ability to work in complicated and unstable environments. Such circumstances are especially noted in the maritime navigation industry due to extensive operations, complicated logistics, and increasing demands to improve cost efficiency and sustainability (Baddar et al., 2025; Ma et al., 2024). Continuous tracking of organizational performance through Key Performance Indicators (KPIs) has become a strategic device for managing operational processes (Alghafes et al., 2024; Misra et al., 2024). Within this context, incentive-based performance systems have emerged as innovative mechanisms to enhance employee engagement, influence behavior, and promote a results-driven culture. These systems incorporate real-time feedback, clear goals, performance recognition, and constructive competition to motivate and guide teams across operational settings.

Research has emphasized the success of motivation models focused on performance in enhancing intrinsic motivation (Yan et al., 2022), task performance (Diefenbach et al., 2024), knowledge retention (Nguyen et al., 2025), and adherence to organizational processes (Diefenbach et al., 2024). Its application spans various domains, including software development (Tolonen et al., 2015), human resource management (Naseer et al., 2023), corporate training (Kim et al., 2021), and sustainable supply chains (Sharafuddin et al., 2022). These models enhance interaction, skill development, and behavior alignment through structured feedback and strategic communication, effectively aligning individual motivations with organizational goals (Yan et al., 2022).

A widely adopted framework for implementing structured performance-based systems typically involves a cyclical process of setting clear objectives, identifying and measuring desired behaviors, understanding target groups, establishing evaluation criteria, providing feedback loops, and recognizing performance outcomes. These steps are consistent with meta-analytic findings on strategic performance measurement systems (Endrikat et al., 2020). Furthermore, these methods have demonstrated encouraging results in multiple fields, such as education (Nie et al., 2023), cybersecurity (Yao et al., 2023), IT task management (Cerroni et al., 2021), and workplace safety (Bęś & Strzałkowski, 2024).

When applied to KPIs, incentive-based performance and motivational models have demonstrated significant contributions to improving financial results, shipping punctuality, productivity, client satisfaction, and compliance indicators (Mtisi, 2022). Their application has also been associated with decreasing downtime (Colombari et al., 2023; Puram et al., 2022), enhancing public health performance (Yan et al., 2022), optimizing production workflows (Fernández-Raga et al., 2023), and supporting operational efficiency in financial institutions (Mtisi, 2022).

Despite advancements, the intersection among motivational performance systems, KPI-based evaluation, and competitiveness remains underexplored, particularly in maritime navigation sectors, where operational performance simultaneously influences safety, productivity, and profitability (Nunes et al., 2024; Endrikat et al., 2020). In particular, limited empirical research has investigated how structured incentive systems integrated with KPI frameworks contribute to strategic execution and sustained competitive advantage.

This study focuses on two areas that have not been explored much in performance management: using data to assess performance and aligning behavior with goals. A data-driven performance evaluation implies a system where decisions are based on real-time, objective information gathered through digital tools. This method improves transparency, holds people accountable, and provides ongoing feedback (Ojha et al., 2024; J. Wang et al., 2022). In this set up, including key performance indicators in a reward system helps shift performance tracking from looking back to taking action ahead of time. It supports proactive management and team autonomy.

Building on this foundation, behavior alignment complements the data-driven approach by focusing on the human dimension of performance. It denotes the process by which individual and collective actions are progressively oriented toward organizational goals through consistent feedback, recognition, and procedural reinforcement. Drawing on motivational theories such as self-determination theory (Van der Hauwaert et al., 2022) and goal-setting theory (Chang et al., 2019), this study conceptualizes behavior alignment as the internalization of safety, compliance, and operational discipline norms through the systematic application of performance indicators.

The convergence of these two mechanisms—data-driven decision-making and behavioral change through incentives—forms the core of the model analyzed in this research. By linking quantifiable performance data to team-level recognition and feedback, the system fosters a culture of continuous improvement and strategic coherence. This dual perspective responds to a gap in empirical studies, which often emphasize what is measured while neglecting how measurement practices influence action and engagement in complex operational environments, such as large-scale shipping operations.

Two bibliographic searches were conducted in the Scopus database to assess the research gap. The first search, with broader terms (“maritime industry” OR “shipping industry” AND “competitiveness” OR “performance” OR “strategic management” AND “innovation” OR “digital transformation”), returned 5174 articles, confirming strong academic interest in competitiveness and innovation in the maritime sector.

A second, more targeted search combining “indicator-based management” with “maritime industry” and “KPI” yielded no results. This outcome does not indicate a complete absence of related studies but rather reflects a terminological gap in how incentive-based performance systems are framed in the literature. Similar approaches are often called motivational design, performance incentives, or behavioral nudges; however, there is a scarcity of research that directly connects these elements to KPI-centered evaluations in port operations. This scarcity highlights a significant empirical and theoretical gap in the literature on indicator-based management in high-risk logistics environments.

The bibliographic search was conducted in Scopus, selected for its broad interdisciplinary coverage, robustness in engineering, transportation, and applied management sciences—domains central to this study. Scopus also offers advanced search operators and transparent citation tracking, ensuring a rigorous and replicable review process. Although Web of Science was also considered, Scopus was prioritized due to its superior coverage of regional and industry-specific journals relevant to Latin American and maritime logistics contexts. This choice is consistent with best practices in applied management research, where Scopus is frequently recognized as the most comprehensive source for multidisciplinary studies (Okoli, 2015).

Indicator-based management, broadly defined as the systematic use of performance indicators to align behaviors with organizational goals, provides a robust theoretical lens for understanding how incentive-based systems shape organizational behavior (Shephard et al., 2015). Far from being a superficial layer of entertainment, indicator-based management in this study refers to a structured system of incentives, feedback, and recognition designed to reinforce desired routines and align team practices with strategic KPIs. Grounded in motivational design principles (Puram et al., 2022), the model integrates general mechanisms—such as constructive competition and reward—with specific features like performance transparency, progress tracking, and public acknowledgement. Together, these elements satisfy fundamental psychological needs for competence, autonomy, and relatedness (Van der Hauwaert et al., 2022).

Indicator-based management is understood here as the structured use of KPIs not only as diagnostic tools but also as behavioral mechanisms. By linking performance monitoring with systematic feedback and recognition, indicator-based management provides transparency, accountability, and continuous reinforcement of desired practices. This approach draws on motivational and organizational theories by showing how quantifiable performance targets can align daily actions with strategic objectives, particularly in complex and high-risk industries such as maritime logistics.

In high-risk operational environments like port logistics, such features transform routine tasks into measurable and meaningful activities, fostering a culture of proactive engagement. This conceptualization underpins the model analyzed in this research, bridging data-driven performance evaluation with behavioral alignment and demonstrating how KPIs function not only as diagnostic tools but also as behavioral levers that promote continuous improvement and shared accountability.

Based on established literature, it is well-supported to expect that incentive-based performance systems can generate positive outcomes in port operations. Research in organizational behavior and logistics indicates that mechanisms such as structured feedback, recognition, and performance transparency significantly enhance compliance, safety, and operational efficiency (Colombari et al., 2023; Puram et al., 2022; Fernández-Raga et al., 2023). When KPIs are aligned with team-level incentives, they act as behavioral levers that promote proactive engagement, reduce reactive patterns, and reinforce adherence to critical operational routines. This theoretical foundation suggests that the incentive model under analysis is likely to strengthen both behavioral alignment and data-driven performance improvement in maritime teams.

The research question guiding this study is: How can an incentive-based performance model, grounded in KPI monitoring, positively influence operational performance and behavioral alignment in Brazilian port operations?

The objective is to evaluate the anticipated positive impact of a behavior alignment model driven by KPIs on the performance of navigation teams in a Brazilian maritime organization. Academically, this research contributes to the literature by integrating performance monitoring, behavioral engagement, and strategic competitiveness in a sector not widely explored from this perspective. The study provides empirical evidence for future research into motivational systems and organizational performance strategies using a robust multivariate methodological approach and real-world data.

From a managerial standpoint, the study addresses growing market demands for tools that effectively link employee engagement with operational efficiency. By reading how KPI-based incentive models affect performance, this work provides a realistic framework to help execute the method, align group behavior, and sell sustainable competitive advantage.

This study employs cluster analysis, the Wilcoxon signed-rank test, the Kruskal–Wallis H test, and exploratory factor analysis to recognize how established overall performance fashions affect operational indicators. These strategies provide a more comprehensive understanding of performance dynamics compared to univariate methods. The research analyzes the results of 145 safety teams operating in 22 Brazilian ports over two years, examining score evolution, standout units, and the latent factors influencing indicator behavior, thus contributing to enhanced port evaluation and management.

Therefore, the research addresses an important relationship between overall performance control and strategic competitiveness inside maritime operations—a website that has remained underrepresented in industrial contexts. Using a structured, incentive-based performance version grounded in KPIs and empirical data, the study proves that such structures can power measurable upgrades in operational consistency, crew behavior, and organizational alignment. The findings contribute to the theoretical advancement of performance governance frameworks and offer practical, actionable insights for managing operations in complex industrial environments.

The remainder of this article is organized as follows. Section 2 outlines the research methodology, describing the multivariate analytical approach adopted, the sample design, and the statistical procedures applied—including clustering, nonparametric testing, and exploratory factor analysis. Section 3 presents the empirical findings from the analysis of ports, states, and regions and compares high- and low-performing units. Section 4 critically discusses the results, interpreting the evidence considering relevant literature and managerial implications. Finally, Section 5 concludes the paper by summarizing the main contributions, recognizing methodological limitations, and proposing future research and model refinement directions.

2. Materials and Methods

This section outlines the methodological steps involved in carrying out the research. This study examined the impact of a structured, incentive-based performance model on the operational indicators of a Brazilian shipping company.

2.1. Research Context and Unit of Analysis

The object of this study is a Brazilian shipping company with operations in 22 ports distributed across 14 states. The unit of analysis is the onboard operational team, totaling 145 teams whose performance was monitored through a structured, incentive-based performance model between 2022 and 2023. Team performance is measured collectively, with individual contributions aggregated into the team’s overall score for each KPI. The teams are named after the port of operation, followed by a sequential number (e.g., Açu 1, Açu 2). Performance data were collected monthly via the company’s digital monitoring system, ensuring objectivity and consistency.

The incentive-based performance model refers to a management approach that uses predefined operational KPIs to align individual and team behaviors with organizational objectives. The model integrates clear performance targets, real-time feedback, recognition mechanisms, and constructive competition to promote adherence to safety protocols, compliance standards, and operational discipline. Each team is periodically evaluated across ten strategic indicators, with results used both to monitor progress and to stimulate engagement by rewarding high performance. By directly linking incentives to measurable outcomes, the model fosters continuous improvement, operational consistency, and strategic alignment within the organization.

To facilitate a comprehensive performance assessment across locations,

Table 1. Overview of the 22 Brazilian ports analyzed, including the number of teams, annual performance scores (2022–2023), state, and regional distribution.

Ports	Teams	Score 2022	Score 2023	State	Region
Açu	11	126.225	124.195	RJ	Southeast
Aracaju	4	46.581	46.888	SE	Northeast
Belém + Vila do Conde	8	95.656	109.542	PA	North
Cabedelo	1	10.997	11.747	PB	Northeast
Fortaleza	8	105.395	100.407	CE	Northeast
Ilhéus	2	23.061	21.785	BA	Northeast
Imbituba	4	46.097	47.894	SC	South
Itajaí	6	75.947	67.617	SC	South
Maceió	2	21.791	23.802	AL	Northeast
Paranaguá	6	65.069	67.772	PR	South
Recife + Ipojuca	6	66.271	69.622	PE	Northeast
Rio de Janeiro	6	70.854	71.768	RJ	Southeast
Rio Grande	4	41.505	48.807	RS	South
Salvador	10	133.432	123.915	BA	Northeast
Santarém	1	14.641	22.788	PA	North
Santos	10	115.917	113.890	SP	Southeast
São Luís	22	249.493	253.353	MA	Northeast
São Sebastião	8	94.574	92.805	SP	Southeast
Sepetiba	8	92.145	95.000	RJ	Southeast
São Francisco do Sul	6	66.752	69.541	SC	South
Trombetas	4	40.899	45.976	PA	North
Vitória	8	90.109	92.779	ES	Southeast

Note: Data aggregated from 145 onboard teams across 22 ports in 14 Brazilian states.

provides an overview of the 22 Brazilian ports analyzed, including the number of teams, annual performance scores (2022–2023), and their respective states and regions. This overview enables the identification of temporal trends and regional disparities, supporting a detailed evaluation of the indicator-based performance system.

Although there is variation in the number of teams per port—ranging from one (e.g., Santarém) to twenty-two (e.g., São Luís)—the unit of analysis remains the individual onboard team. All statistical analyses (cluster analysis, Wilcoxon signed-rank test, and exploratory factor analysis) were conducted at the team-month level, ensuring that each team contributes equally to the results, regardless of port size or operational scale.

The scores shown in Table 1 represent the annual average performance of each port, calculated from the monthly KPI evaluations of its onboard teams. Each KPI was assessed monthly, and individual team scores were aggregated into an annualized value on a scale from 0 to 1000 points. While Table 1 illustrates the distribution of performance across ports, the statistical analyses were not based solely on these aggregated totals. Instead, the study employed standardized, team-month level data as the primary input for cluster analysis, the Wilcoxon signed-rank test, and exploratory factor analysis (EFA), ensuring methodological robustness and preventing bias from port size or score variability.

The variation in the number of teams per port reflects differences in operational scale, with larger ports operating more teams than smaller ones. To ensure comparability, port-level performance values were calculated as the average of monthly team-level scores. This standardized approach enables meaningful comparisons across ports of different sizes while preserving within-port variability.

This distribution enables a comparative analysis across regions and operational scales, supporting the identification of performance patterns and contextual influences.

2.2. Research Design

This study adopts a single-case study design with a multi-site structure, focusing on one Brazilian shipping company operating across 22 ports in 14 states. The unit of analysis is the onboard operational team, totaling 145 teams monitored between 2022 and 2023. This approach allows for a deep, contextually grounded analysis of how a standardized incentive-based performance model is implemented and performs across different operational environments within the same organizational framework.

The research employs a quantitative, field-based methodology to evaluate the impact of the incentive-based performance model. Data were collected monthly through the company’s digital monitoring system, ensuring objectivity, transparency, and continuity.

Performance is operationalized through ten strategically selected Key Performance Indicators (KPIs), grouped into three core dimensions: Operational Safety, Technical Availability, and Procedural Compliance. These indicators are quantitatively measured, systematically recorded, and directly linked to the model’s feedback and recognition mechanisms, enabling a data-driven evaluation process.

To address the research objectives, a multi-method analytical strategy was applied:

• The Wilcoxon signed-rank test was used to assess significant performance changes over time (2022 vs. 2023).
• The Kruskal–Wallis test identified differences in performance across ports and states.
• Cluster analysis grouped teams into performance tiers (high, intermediate, low) based on similarity in KPI scores.
• Exploratory factor analysis (EFA) evaluated the stability of the underlying factor structure across time, confirming the construct validity and robustness of the measurement model.

This integrated approach ensures both statistical rigor and practical relevance, allowing for a comprehensive assessment of how the incentive-based system influences performance dynamics and organizational alignment.

2.3. Indicators

Ten operational indicators, reflecting safety, compliance, operational discipline, and technical maintenance practices. These are grouped into three core dimensions:

• Operational Safety: WS Guardian, Onboard Inspection, Cable Break, Emergency Drill;
• Technical Availability: Planned Repair Execution, Certificate Renewal, Personnel Documentation;
• Procedural Compliance: Safety Dialogue, SMS Meeting, Behavior Observation.

This three-factor structure was empirically validated through exploratory factor analysis (EFA) and was identified and summarized in

Table 2. KPIs and their operationalization.

Indicator	Type	Description/Measurement
WS Guardian	Safety practice	Number of at-risk behaviors reported and registered through the WS Guardian digital platform (monthly count).
Safety Dialogue	Safety culture	Frequency of structured safety conversations held between team leaders and crew (number per month).
Emergency Drill	Safety readiness	Number of emergency response simulations performed by teams (monthly).
SMS Meeting	Procedural compliance	Monthly meetings held to review the Safety Management System (binary: held = 1; not held = 0).
Technical Availability	Operational efficiency	Percentage of time vessels were technically available for operations (ratio of available hours to total hours).
Certification Renewal	Compliance	Proportion of crew certificates renewed within validity deadlines (percentage per month).
Near Miss Reporting	Preventive action	Number of near-miss incidents formally reported by crew members (monthly count).
Operational Safety	Composite indicator	Aggregated score of safety-related practices (weighted average of WS Guardian, Safety Dialogue, Emergency Drill, and SMS Meeting).
Procedural Compliance	Compliance culture	Percentage of adherence to mandatory safety and operational procedures (monthly score).
Overall KPI Score	Performance outcome	The final performance index combines all KPIs and is used for team ranking and incentive allocation.

Note: All KPIs are quantitative measures obtained from the company’s digital monitoring system.

.

Each indicator aligns with the company’s operational excellence model, which promotes individual and team performance in navigation environments. Scores were computed monthly, with each indicator contributing up to 1000 points, recalculated according to the specific rules established for each metric. Each team collected data digitally and consolidated it through a centralized system for monthly managerial evaluation.

In the incentive-based performance model adopted in this study, team evaluation and rewards are based on ten operational KPIs, which serve as the core indicators for monitoring performance and determining incentive allocation. These KPIs were selected based on their strategic relevance to port operations and defined through a collaborative process involving operational leaders and safety managers, ensuring alignment with both organizational objectives and frontline realities. Structured around three core dimensions—operational safety, technical availability, and procedural compliance—the indicators collectively reflect the multifaceted nature of port performance.

The system functions using a data-driven approach: every KPI is assessed objectively through digital records or recorded activities, reducing subjectivity and allowing for transparent, real-time monitoring of performance. At the same time, the indicators are designed to promote behavior alignment by reinforcing desired routines—such as proactive risk reporting, preventive maintenance, and regular safety discussions—through monthly feedback and team recognition—mechanisms aligned with motivational design principles that foster habit formation and collective accountability. In this way, the KPIs serve simultaneously as performance measures and behavioral drivers, ensuring that evaluation is not only retrospective but also actionable.

These indicators serve as both diagnostic tools and behavioral levers, providing quantifiable data for evaluation while encouraging the adoption of safety and compliance routines across teams. All ten KPIs were systematically recorded through the company’s digital monitoring system on a monthly basis for each of the 145 onboard teams operating in 22 Brazilian ports during 2022 and 2023. The KPIs serve as the dependent variables in the analysis, enabling a quantitative assessment of performance dynamics over time. This approach helps us see performance not just as a general concept, but as real and consistent actions that are directly connected to the incentive system.

The scoring system was graded on a scale from 0 to 1000 points, reflecting the highest potential performance a team could achieve. Collecting data for each person allowed for a close look at how each team did, and combining the results helped find the average performance score for each region in Brazil. This regional segmentation facilitated the identification of percentage increases in performance between 2022 and 2023 and was complemented by monthly trend analyses to examine score evolution at both port and regional levels.

This dual function—measurement and motivation—ensures that the evaluation process is not only statistically robust but also organizationally transformative, fostering a culture of continuous improvement grounded in evidence and shared responsibility.

The following section outlines the unique contribution of each statistical technique employed to ensure methodological transparency and clarify the rationale for each analytical choice.

2.4. Statistical Methods

To ensure methodological transparency and to clarify the rationale for each analytical choice, this section provides a detailed explanation of the unique contribution of each statistical technique employed.

Cluster analysis (K-means) was used to identify natural groupings of teams based on performance similarity, enabling the identification of distinct performance tiers (high, intermediate, and low). This unsupervised technique supports strategic decision-making by highlighting best practices in high-performing units and areas requiring intervention in low-performing ones.

The Wilcoxon signed-rank test was selected to assess significant changes in performance over time (2022 vs. 2023) at the regional level. As a nonparametric alternative to the paired t-test, it is robust to nonnormal data distributions and outliers—common characteristics in operational performance datasets—making it ideal for evaluating longitudinal impacts.

The Kruskal–Wallis H test was used to compare performance differences across independent groups such as ports and states. As a nonparametric alternative to one-way ANOVA, it does not require assumptions of normality or homogeneity of variances, making it highly suitable for real-world operational data where score distributions vary significantly between units.

Exploratory factor analysis (EFA) was employed to test the stability of the underlying dimensions across time. By conducting separate EFAs for 2022 and 2023, the study evaluated the temporal consistency of the factor structure, confirming the construct validity and robustness of the measurement model.

This multi-method strategy ensured both robustness and replicability of results, providing a clear basis for linking performance outcomes to the incentive-based model.

2.5. Regional Analysis

To assess regional disparities in operational performance, the research segmented the dataset into four macro-regions of Brazil: North, Northeast, Southeast, and South. This categorization allows for an aggregated view of performance patterns across diverse operational and socio-economic contexts. For each region, average scores were calculated for 2022 and 2023, and the evolution of these scores was tracked over time.

To evaluate whether there were statistically significant differences in performance scores among these regions, the Kruskal–Wallis H test was applied. This nonparametric method is suitable for comparing three or more independent groups when the assumption of normal distribution is violated (Woźniak & Wereda, 2023), which was the case for the performance scores across the regions. Unlike ANOVA, the Kruskal–Wallis test does not require equal variances or normality (Pasaco-González et al., 2023), making it an appropriate choice for this study’s dataset. In this context, the test assessed whether the median performance scores differed significantly among the North, Northeast, Southeast, and South regions, providing a robust basis for detecting regional disparities in operational outcomes.

The Kruskal–Wallis test results showed that regional differences were not uniform. Only the South region demonstrated a statistically significant performance improvement from 2022 to 2023 (Z = −2.434; p = 0.015), suggesting that region-specific factors—such as operational maturity, leadership style, and institutional support—may influence the adoption and effectiveness of performance management systems. These findings underscore the importance of tailoring strategies to local contexts to optimize results.

2.6. Team Segmentation and Temporal Analysis

Developed in the early 1960s, the K-Means algorithm is widely recognized in the literature for its simplicity, computational efficiency, and extensive use in non-hierarchical clustering tasks (Aljabbouli et al., 2020; Carretillo Moctezuma et al., 2025; Sari et al., 2018). It employs an objective function that minimizes within-cluster variance, thus characterizing the clustering process as an optimization problem.

The clustering techniques of k-means divided the teams into groups based on performance (Ehrgott et al., 2018). This approach sorted the 145 teams into three groups: low, medium, and high performance, depending on their average score from 2023. At the same time, the 20 teams with the lowest scores in February 2022, when the incentive-based performance model was first implemented, were selected, and the evolution of each indicator throughout 2023 was analyzed. The average evolution of these teams’ indicators was also calculated, comparing the results for 2022 and 2023. This analysis enabled the identification of specific improvements and highlighted the indicators that had the most significant impact on the performance of these teams.

2.7. Factor Analysis of Performance Indicators

Additional statistical analyses were conducted to understand the performance indicators’ structure better and ensure that the applied methods were robust and reliable. An Exploratory Factor Analysis (EFA) was employed to identify latent factors influencing team performance based on the ten structured performance indicators (Khan et al., 2017; Weiner et al., 2017). The Kaiser–Meyer–Olkin (KMO) measure of sampling adequacy and Bartlett’s test of sphericity were used to assess the suitability of the dataset for factor extraction (Costales et al., 2022). Based on eigenvalues and inspection of the scree plot, three main dimensions were consistently identified in 2022 and 2023: Operational Safety, Technical Availability, and Procedural Compliance. Together, these factors explained 74% of the total variance in 2022 and 76% in 2023, demonstrating strong explanatory power and temporal consistency.

Additionally, nonparametric tests were used to make sure the results were valid at different levels of analysis:

• The Wilcoxon signed-rank test was selected because it is a nonparametric alternative to the paired t-test. It is suitable for analyzing paired samples—in this case, performance scores for the same ports and regions in 2022 and 2023—when the assumption of normality is violated. Its robustness to outliers and suitability for ordinal or nonnormally distributed data made it appropriate for this study’s dataset.
• The Kruskal–Wallis H test was employed as a nonparametric alternative to one-way ANOVA. It allows the comparison of three or more independent groups (e.g., ports, states, or regions) without assuming a normal distribution or equal variances. This test is particularly valuable in operational performance datasets where score distributions may vary substantially among groups, ensuring the valid detection of statistically significant differences.
• The K-means clustering method was used to divide teams into three groups—high, medium, and low performance—based on their scores from the 2023 annual assessment. This grouping helped with further comparison and figuring out what works best.

These methods provided a whole way to examine how performance changed over time, find common patterns in the structure, and check differences between groups without assuming data followed a standard curve. Using factorial and cluster methods together allowed a better understanding of how the incentive-based model—conceptually grounded in indicators-based management principles—shaped behavioral patterns and operational outcomes across different work settings. In this context, indicators-based management refers to the use of structured feedback, team recognition, and constructive competition to transform routine operational tasks into measurable, engaging, and goal-oriented activities (Miyamoto & Kiyota, 2017).

2.8. Summary of Methods and Analytical Relevance

This study used a multivariate approach to check how well a structured performance model works for operational teams in Brazilian ports. The research looked at three key areas: sorting out different performance levels, looking at results over time, and checking if the evaluation system is set up correctly.

First, K-means clustering was used to group ports, states, and regions into high-, medium-, and low-performance categories based on average KPI scores, year-over-year variation, and the statistical significance of change. Second, nonparametric statistical tests were applied. The Wilcoxon signed-rank test assessed paired performance data between 2022 and 2023, while the Kruskal–Wallis test compared performance across clusters. Third, an Exploratory Factor Analysis (EFA) was performed separately for each year to validate the internal structure of the ten KPIs, confirming three consistent factors across both years: Operational Safety, Technical Availability, and Procedural Compliance.

As a result, the elevated values of the Kaiser-Meyer-Olkin (KMO) measure, coupled with the significant outcomes from Bartlett’s test, indicated that the model demonstrated substantial statistical improvements, suggesting a strong interconnection among the variables. This finding implied that the data was well-prepared for more complex analyses, enabling the exploration of deeper insights. The thorough methodology utilized in this study not only enhanced the reliability of the analysis but also addressed and resolved several longstanding issues associated with the use of individual scores.

3. Results

3.1. Cluster Analysis by Port

A K-means clustering analysis was performed to classify the 22 ports into three performance groups: high performance, intermediate performance, and low performance. The classification was based on each port’s average annual scores across the structured performance indicators.

Table 3. Port performance clustering: classification based on 2023 average scores.

Cluster	Ports	Teams	Observations
High performance	Maceió, Rio de Janeiro, Imbituba, São Francisco do Sul, Sepetiba, Vitória	58	Low average gap in relation to the maximum score (below 30 points); good evolution in 2023
Intermediate performance	Santos, Paranaguá, Itajaí, Recife + Ipojuca, Ilhéus, Cabedelo, Belém + Vila do Conde, Salvador, Trombetas	80	Moderate variation between indicators; some showed significant improvement in 2023
Low performance	Açu, Belém + Vila do Conde, Trombetas, Santarém, Ilhéus, Rio Grande	7	Larger gaps in relation to the target, some with worsening in 2023

presents this classification into clusters.

Table 3 shows how each port is connected to its teams and the groups found in the analysis. Ports in the high-performance group are those where the teams obtained the highest scores on the performance indicators. High-performance ports consistently showed smaller score gaps relative to the maximum possible score (below 30 points) and demonstrated improvement in 2023. In contrast, low-performance ports had larger gaps and inconsistent team behavior, with some showing deterioration over time.

The nonparametric Kruskal–Wallis test was applied to determine whether there were statistically significant differences among the ports within the same year of analysis. This test is appropriate for comparing independent groups without assuming a normal distribution. When necessary, multiple comparisons (post hoc) were conducted to identify specific differences between the groups. The results showed that in both 2022 and 2023, there were significant differences in the average port scores (p < 0.001 in both years), which suggests that the performance of different port regions varied within the organization being studied.

The data was also split by port so we could examine it in more detail. Figure 1 shows the average scores from each port team in 2022 and 2023, making it easier to compare how performance changed over time. This in-depth method helps us understand the differences between regions and how the metrics developed during the studied period.

The Wilcoxon signed-rank test was used to examine how the average performance of the ports changed from 2022 to 2023. This test provides a statistical way to understand how operational results change over time. It was a good choice because the performance scores were in order, but did not follow a normal distribution.

The analysis revealed that five ports—Imbituba, Aracaju, Fortaleza, São Luís, and Maceió—experienced statistically significant improvements in their average scores over the two years. The improvements were evident in areas connected to operational safety, showing that the organized performance model helped these places follow. The findings underscore the remarkable potential of thoughtfully crafted incentive frameworks to foster significant behavioral enhancements, particularly when underpinned by established routines and driven by proactive leadership engagement.

3.2. Cluster Analysis by State

The 14 states with evaluated ports were checked using their average indicator scores from 2022 and 2023, the percentage change year after year, and whether the changes were statistically significant, which was tested using the Wilcoxon test. Based on these factors, the states were divided into three performance groups:

• High performance (average ≥ 965 or variation ≥ 5% with p < 0.05),
• Intermediate (average between 930 and 964 or variation between 2 and 5%), and
• Low performance (average < 930 and variation < 2% or declining trend).

This stratification helped understand the states’ maturity in their operations and how much the structured performance model affected different organizations and regions. It made it easier to spot differences in how engaged teams were, how well they followed processes, and how quickly they responded to the analysis based on indicators. By segmenting the states into distinct performance clusters, the analysis supported a more targeted understanding of which areas benefited most from the intervention and where structural or cultural barriers may still exist.

Table 4. Cluster-based classification of Brazilian states by operational performance (2022–2023).

Cluster	States	Score 2022	Score 2023	Δ%	Teams	Z Statistic and p-Value
High performance	AL	908	992	9.20%	2	−2.511, 0.012
	CE	954	996	4.40%	8	−2.040, 0.041
	PB	916	979	6.80%	1	−1.867, 0.062
	PE	920	967	5.10%	6	−1.961, 0.050
	MA	893	960	7.40%	22	−2.786, 0.005
	RS	865	948	9.60%	4	−2.824, 0.005
Intermediate performance	SE	970	977	0.70%	4	−0.400, 0.689
	RJ	964	970	0.60%	25	−0.981, 0.327
	SC	929	964	3.70%	16	−1.256, 0.209
	ES	939	966	3.00%	8	−0.785, 0.433
	SP	927	950	2.50%	18	−0.549, 0.583
	PR	904	941	4.10%	6	−1.256, 0.209
Low performance	PA	886	931	5.10%	13	−1.490, 0.136
	BA	942	908	−3.6%	12	−1.255, 0.209

organizes the 14 states into distinct clusters based on their performance levels, categorizing them as high, intermediate, and low. This classification not only highlights the varying degrees of effectiveness and maturity in the implementation of the structured performance model but also provides a valuable comparative perspective across these diverse units. By examining these clusters, we can gain deeper insights into the strengths and challenges faced by each state in its pursuit of enhanced performance.

The following analysis examines the average score results by state for 2022 and 2023, as detailed in Figure 2. This step enables a more granular view of performance evolution, considering the specific characteristics of each federal unit where the company operates.

Five of the 14 states showed statistically significant improvement (p < 0.05) in the comparison between 2022 and 2023, according to the Wilcoxon test for paired samples: Alagoas (AL), Ceará (CE), Paraíba (PB), Pernambuco (PE), Maranhão (MA), and Rio Grande do Sul (RS). The results show that using incentives to improve performance helped these areas, leading to noticeable improvements in how well operations are running and showing that local teams were more involved. Even though most states saw progress, Bahia (BA) worsened, suggesting particular issues related to the local situation or how the organization was set up.

3.3. Cluster Analysis by Region

A regional analysis was conducted across Brazil’s four major geographic regions: the South, Southeast, Northeast, and North. The regions were grouped based on four key areas for analysis:

• Average aggregated indicator scores for 2022 and 2023 were a representation of the absolute level of performance.
• Year-over-year percentage change is used to capture progress trends.
• Statistical significance of changes (Wilcoxon test with p < 0.05)
• Internal consistency of regional results, measured by standard deviation and score amplitude

Consolidating these criteria also allowed the formation of three clusters, reflecting differing levels of operational maturity.

Table 5. Average regional scores of operational teams in 2022 and 2023 with cluster assignment.

Cluster	Region	Score 2022	Score 2023	Δ%	Teams	Z Statistic and p-Value
High performance	South	905	949	+4.8%	20	−2.434, 0.015
Intermediate performance	Northeast	925	960	+3.8%	43	−1.334, 0.182
	Southeast	947	962	+1.6%	51	−0.196, 0.844
Low performance	North	886	931	+5.1%	31	−1.490, 0.136

presents the average scores of operational teams by region and their corresponding cluster classification, highlighting regional disparities and levels of performance consistency. The column represented by the Δ% denotes the percentage variation in average scores between 2022 and 2023, providing insight into performance changes over time.

The Wilcoxon test was performed to verify whether there was a statistically significant change in regional averages between 2022 and 2023. Even though all regions increased their average scores, only the South region indicated a statistically significant improvement (p = 0.015), increasing from 905 in 2022 to 949 in 2023 (+4.8%).

Additionally, the Z statistic of −2.434 suggests that the improvement did not occur by chance and that the structured performance model had a measurable impact in this region, predominantly on the WS Guardian, SMS Meeting, and Emergency Drill indicators.

The Southeast region had the highest average score in 2023, but the difference between it and the Northeast was not very big. This pattern indicates that the model fostered performance gains even in previously high-performing regions, as illustrated in Figure 3.

The other regions did not show statistically significant differences or exhibited only minor fluctuations (p > 0.05), which can be attributed to three main factors: (i) already high scores in 2022, limiting the potential for further improvement (ceiling effect); (ii) operational heterogeneity among ports, which increases internal variability; and (iii) lower adherence or maturity of the structured performance model, which still requires time to be consolidated. The analysis shows that improving skills, ensuring the plan matches goals, and adapting to local situations can improve the model in those areas.

Although the North region showed growth (+5.1%), the difference was not statistically significant, indicating that contextual, cultural, and structural factors may still limit the impact of the performance model in this area. To maximize its benefits, it is recommended that the model be customized to local needs, leadership engagement be reinforced, and team implementation monitoring be intensified.

3.4. Exploratory Factor Analysis of Indicators

An Exploratory Factor Analysis (EFA) was performed separately for 2022 and 2023 to assess the structural consistency of the ten operational performance indicators. The results showed that the data were good enough for this type of analysis. In 2022, the Kaiser-Meyer-Olkin measure was 0.82; in 2023, it was 0.85. Both years also had significant Bartlett’s sphericity tests, with p-values less than 0.001.

Three latent dimensions were identified in both years:

• Factor 1—Operational Safety: WS Guardian, Onboard Inspection, Cable Break, Emergency Drill.
• Factor 2—Technical Availability: Planned Repair Execution, Certificate Renewal, Personnel Documentation.
• Factor 3—Procedural Compliance: Safety Dialogue, SMS Meeting, Behavior Observation.

Together, these factors explained 74% of the variance in 2022 and 76% in 2023. Factor loadings increased over time, particularly for Operational Safety and Procedural Compliance, indicating improved clarity and internalization of the evaluation criteria by the teams.

The factor structure’s consistency supports the original constructs’ robustness and validates the indicators’ theoretical grouping. However, Behavior Observation persistently showed lower loadings, possibly due to the subjective nature or inconsistent data reporting.

EFA thus proved to be an effective tool for validating the model’s dimensional coherence. It is recommended that these indicators be continued, along with regular monitoring. Future refinements include testing for discriminant validity, assessing structural invariance across regions or port types, and integrating additional indicators, such as communication practices, safety climate, or nonconformity resolution.

3.5. Comparative Analysis: High vs. Low-Performance Ports

The comparison between high- and low-performing ports revealed clear performance disparities across the ten indicators. High-performing ports—such as Imbituba, Maceió, São Luís, Aracaju, and Sepetiba—maintained average scores above 900, exhibited low internal variability, and showed consistent improvement across most indicators. In contrast, ports like Trombetas, Santarém, and Açu presented average scores below 500, irregular performance, and little measurable progress.

Indicators such as WS Guardian, Emergency Drill, and SMS Meeting showed the most significant performance gaps between the two groups. For instance, the average score for WS Guardian reached 970 in high-performing ports versus 274 in low-performing ones—suggesting stark differences in preventive engagement and operational maturity.

High performers consistently followed structured routines and collaborative practices, such as meetings, feedback, and documentation. In contrast, low performers displayed reactive behavior, minimal participation, and resistance to the structured performance model.

Table 6. Comparison of indicator performance between high- and low-performing ports.

Evaluated Criteria	High Performance Ports	Low Performance Ports
Strong indicators	WS Guardian, Emergency Drill, Planned Repair, SMS Meeting	None consistently strong
Weak indicators	Punctual and with good recovery	WS Guardian, Emergency Drill, SMS Meeting
Standard deviation between teams	Low (consistent teams)	High (internal inconsistency)
Adherence to goals	High (score close to 1000)	Low (averages < 500 in several indicators)
Predominant type of action	Preventive, planned, and continuous	Reactive or absent
Collective practices	Consolidated (meetings, training, feedback)	Resistance to adhesion and low perceived value

presents the most prominent performance differences between high- and low-performing port indicators, listed by their operational names as defined in Section 2.1.

As previously highlighted, specific indicators—such as WS Guardian, Emergency Drill, and SMS Meeting—appear in both columns due to their high discriminatory power. Their inclusion underscores their pivotal role in differentiating operational maturity levels between ports. High-performing ports consistently excelled in these areas, while low-performing ones struggled to meet baseline expectations. This dual presence reinforces the diagnostic value of such indicators for recognizing excellence and identifying areas needing targeted intervention.

These contrasts highlight a key limitation of the current model: the equal weighting of all ten indicators, regardless of their strategic relevance or implementation complexity. Future model iterations could incorporate differentiated weights using multi-criteria decision-making approaches such as AHP or TOPSIS to address this.

It is also suggested that qualitative field assessments be conducted in ports that are not performing well to identify organizational or cultural issues. Structured benchmarking with top-performing ports can help spread good practices.

Finally, robust statistical techniques—Wilcoxon, Kruskal–Wallis, K-means, and EFA—enhanced the findings’ reliability and provided actionable insights for decision-making, prioritization, and strategic planning.

4. Discussion

This investigation presents compelling evidence that when used wisely in work settings, systems that reward performance in a structured way can significantly boost team results and help keep the organization aligned. Using a multivariate analytical approach that includes clustering, nonparametric testing, and exploratory factor analysis, the study goes beyond the limitations of simple scoring methods, providing a better understanding of how performance indicators change in complex organizational settings. The analysis used data from 22 ports in 14 Brazilian states between 2022 and 2023 and found that performance kept improving, especially in the southern part of the country, where the improvements were statistically significant.

The stability of the underlying factor structure across the ten indicators reinforces the robustness of the model and its applicability for ongoing performance management. Cluster analysis effectively distinguished high-, intermediate-, and low-performing units at both port and state levels, emphasizing the need for context-sensitive strategies—especially in underperforming regions such as the North. The model worked well in encouraging participation and matching behavior. However, the results are different for each team and each measure, which means there is room for improvement and the need for more customized management approaches.

4.1. Summary of Key Findings and Implications

The investigation shows that reward systems based on performance can help improve operations and make behavior more consistent. Cluster analysis revealed precise segmentation into three performance tiers across ports and states, and was found among ports and states by grouping data. Ports that did well, like Maceió, Imbituba, Sepetiba, and Vitória, had slight differences in their performance and showed significant improvements in 2023.

In contrast, ports such as Trombetas, Santarém, and Açu showed varying or worse performance with more unpredictable results. At the regional level, only the South showed a clear improvement that was statistically significant (Wilcoxon Z = −2.434; p = 0.015), showing better use of organized methods to improve performance. States like Alagoas improved by 9.2% and Ceará by 4.4%, but Bahia saw a drop of 3.6%.

These differences show how important local leaders, the culture of the organization, and how well the team works together are in affecting performance results.

The observed performance improvements—particularly in indicators such as WS Guardian, Emergency Drill, and SMS Meeting—transcend mere statistical significance, offering behavioral evidence of organizational change. These signs show that we need to take early, teamwork-based, and preventative steps, like pointing out dangerous actions, practicing what to do in emergencies, and having regular safety discussions. Their consistent improvement suggests a shift from reactive compliance to internalized safety routines that reflect deeper behavioral alignment.

This change matches ideas from indicators-based management and motivational design (Puram et al., 2022), where clear feedback and praise help form good habits by encouraging the right actions over time.

The system’s reliance on real-time performance metrics is central to this transition. Data were digitally collected, aggregated, and displayed to teams in real time through scoreboards and monthly reviews. Such transparency converts abstract goals into tangible outcomes, enabling teams to monitor progress, identify gaps, and adjust their actions accordingly. The use of cluster analysis, Wilcoxon tests, and exploratory factor analysis enhances analytical rigor, moving beyond anecdotal assessment toward statistically validated insights that inform managerial decision-making.

Finally, the stability of the factor structure—Operational Safety, Technical Availability, and Procedural Compliance—over time indicates that teams are not merely responding to incentives but are developing a shared understanding of what performance entails. This cognitive alignment, reinforced by continuous data feedback, reflects the emergence of a more mature organizational culture within the Brazilian shipping company, in which evidence rather than ad hoc directives increasingly guides behavior. These results show that managers should focus on systems that do more than measure performance. They should also encourage good behavior by providing clear feedback and making sure team members hold each other accountable. Overall, these ideas highlight how combining performance tracking with behavior support can help create positive changes in workplace culture, especially in high-risk areas like those found in growing markets such as Brazil.

4.2. Structural Validity of the Indicators

The three-factor structure that emerged from the EFA reflects distinct but interrelated dimensions of port team performance:

• Operational Safety captures proactive risk management through indicators like WS Guardian (peer reporting of unsafe acts) and Emergency Drill (preparedness for critical events). These require collective vigilance and preventive behavior, making them strong proxies for behavior alignment.
• Technical Availability reflects the reliability of equipment and documentation, with indicators such as Planned Repair Execution and Certificate Renewal. These are data-intensive and process-driven, exemplifying data-driven operational control.
• Procedural compliance measures adherence to organizational routines, including safety dialogue and SMS meetings. Their improvement indicates increased internalization of formal processes, a key aspect of cultural and behavioral alignment.

These three dimensions explained 74% of the variation in 2022 and 76% in 2023, showing that the results were consistent and stable over time. The strength of each dimension improved, especially for Operational Safety and Procedural Compliance, which suggests people became clearer and more confident about what is being measured. However, the Behavior Observation part had lower scores because it is harder to measure; people do not report it the same way every time, or it does not happen very often, so there may be room to improve the model.

4.3. Practical Contributions and Management Insights

This study offers several practical contributions. By demonstrating that structured KPI-based incentives improved indicators such as safety meetings, emergency drills, and risk reporting, incentive systems can effectively embed preventive routines into daily operations. These findings are particularly relevant in high-risk industries, where compliance failures can generate severe operational, financial, and reputational costs (Fernández-Raga et al., 2023). The model provides a replicable template for organizations seeking to combine performance monitoring with behavioral reinforcement, extending its applicability beyond the port sector to logistics, energy, and manufacturing operations (Colombari et al., 2023). In practice, the integration of real-time KPI monitoring with incentive-based feedback reduces downtime, increases accountability, and supports continuous operational learning.

From a managerial perspective, the results highlight that performance indicators should not be treated solely as diagnostic measures but also as behavioral levers. Real-time feedback and recognition mechanisms transform KPIs into instruments for motivation, accountability, and peer learning. Managers can leverage these insights to design targeted interventions for low-performing teams, allocate resources more strategically, and cultivate a proactive safety culture. These findings echo prior evidence showing that indicators-based management systems enhance engagement, self-regulation, and cooperation in organizational contexts (Puram et al., 2022; Colombari et al., 2023). Notably, the stability of the factor structure over time suggests that teams are developing a shared understanding of performance expectations, reinforcing the role of managers as facilitators of cultural alignment rather than enforcers of compliance.

4.4. Contributions to the Literature

Theoretically, this research contributes to performance management literature by bridging data-driven evaluation with behavioral alignment in a real-world logistics setting. Performance management frameworks have long emphasized the need to link measurement with strategy (Neely et al., 2005; Franco-Santos & Otley, 2018). However, few empirical studies have operationalized this connection in high-risk operational environments. By conceptualizing indicators-based management as a structured incentive model and empirically validating its effects through multivariate statistical methods, the study enriches both indicators-based management research (Colombari et al., 2023; Puram et al., 2022) and maritime management literature (Shephard et al., 2015). In doing so, it positions the research at the intersection of organizational behavior, performance management, and transportation studies. It opens avenues for comparative studies in other industries and emerging markets where incentive alignment remains underexplored.

4.5. Limitations and Future Research

Despite its contributions, the study has a few limitations:

• The equal weighting of indicators may obscure their varying strategic importance. Future research should explore dynamic weighting using MCDM techniques, potentially incorporating uncertainty measures.
• The reliance on quantitative data alone limits insight into the motivational mechanisms behind performance changes. Mixed-methods studies, including interviews or focus groups, could provide richer interpretations.
• Although appropriate for nonnormal data, nonparametric methods may limit the detection of subtle interaction effects.
• The findings are context-specific to the port industry and may not generalize to other sectors without adaptation.

Given the complexity of performance management in operational environments, future studies should consider broadening the analytical framework to incorporate quantitative and qualitative dimensions. Such a combination may involve combining statistical modeling with field observations or interviews to understand behavioral drivers and barriers better. Additionally, researchers could explore:

• Implementing dynamic indicator weighting using methods like AHP or TOPSIS, to refine version responsiveness, which may include confidence intervals (Sellitto et al., 2022);
• Conducting longitudinal assessments over extended periods to evaluate sustained impacts;
• Employing qualitative techniques to explore user perceptions and cultural receptivity;
• Designing targeted interventions for low-performing units (e.g., behavior nudges, peer benchmarking);
• Integrating the performance model into digital platforms to enhance feedback and user engagement.

By addressing these directions, future research can deepen the understanding of how structured performance models contribute to continuous improvement, behavior alignment, and organizational transformation in complex operational systems.

5. Conclusions

From a theoretical standpoint, this study contributes to performance management literature by demonstrating how data-driven evaluation and behavior alignment can be operationalized and empirically validated in a high-risk, complex logistics environment. By applying multivariate statistical methods—clustering, nonparametric testing, and exploratory factor analysis—the research ensures that assessment is not only systematic but also rigorous and replicable. The stability of the factor structure over time confirms that the indicator-based system functions as a coherent, data-informed framework capable of guiding long-term improvement. This finding is consistent with broader evidence highlighting the role of integrated measurement systems in driving strategic alignment and organizational learning (Neely et al., 2005; Franco-Santos & Otley, 2018). Moreover, this study is among the few to empirically validate a behavioral performance model using advanced analytics in a real-world logistics setting.

Practically, the findings show that when performance data were embedded within a culture of recognition and feedback, they become powerful tools for shaping behavior. High-performing ports exhibit not only higher scores but also consistent, preventive, and team-focused practices, demonstrating that the incentive model effectively aligned individual action with organizational objectives. These results align with prior studies showing that indicator-based management and structured feedback mechanisms enhance engagement, compliance, and proactive behavior in organizational contexts (Colombari et al., 2023; Puram et al., 2022). Conversely, the reactive patterns observed in low-performing ports underscore the importance of contextualized managerial interventions that address both technical and cultural barriers to alignment, echoing findings in other safety-critical industries (Fernández-Raga et al., 2023).

Taken together, these insights suggest that incentive-based performance systems, when supported by transparent KPI monitoring and structured recognition, have the potential not only to improve immediate operational results but also to foster long-term cultural transformation in organizations. Such a transformation is particularly relevant for high-risk industries in emerging markets, where the integration of behavioral incentives with performance monitoring remains underexplored but holds considerable promise for advancing both efficiency and safety.

However, some limitations must be acknowledged. The equal weighting of indicators may overlook strategic differences in relevance, and the exclusive reliance on quantitative data limits understanding of motivational drivers and organizational dynamics. Future research should explore multi-criteria decision-making methods (e.g., AHP or TOPSIS), incorporate qualitative methods to capture team engagement, and assess the model’s generalizability across other sectors.