Period: November 2024 – January 2025
Port of Antwerp-Bruges (PoAB) is a merged organization that manages the ports of Antwerp and Zeebrugge and employs more than 1,700 people in a variety of operational and support roles.

In the fall of 2024, HR noticed some notable trends in the number of overtime hours. These spikes deviated from the normal seasonal pattern and appeared to be increasing in certain roles, months, and teams. Starfish Consultancy was brought in to analyze these trends, identify contributing factors, and formulate concrete recommendations.

The goal of the project was threefold:

• To provide insight into overtime trends (at the annual, monthly, and employee levels);
• To identify the main causes and at-risk groups;
• To translate these insights into targeted actions that support company policy.

The project proceeded in three major phases, with weekly coordination meetings held with an internal steering committee consisting of HR staff and, where relevant, experts and end users of the data.

Phase 1 – Data Exploration and Desk Research
We collected and documented data from their scheduling tool, including shift information, absences, workplace accidents, training hours, and understaffing data. This data was aggregated at the monthly level and linked to employee characteristics (center, position, age, tenure). In parallel, we conducted desk research on external factors such as seasonal illness.

Phase 2 – Analyzing Data
In this phase, we identified the trends and causes of overtime through analyses in RStudio. The analyses were performed on-premises on a Port of Antwerp-Bruges company laptop to ensure data security.

This phase went through multiple iterations, during which interim results were reviewed with the internal steering committee to assess their validity and contextualize them within the organization. This feedback was then incorporated to adjust subsequent analyses—and, in some cases, the scope of the project.

Descriptive Analyses (What Are We Seeing Happen?)
We examined how overtime evolved by month, by position, by department, and by employee type. This allowed us to identify peak periods and uncover unusual patterns. Additionally, we examined differences at the employee level, such as the proportion of employees who regularly work long shifts.

Diagnostic analyses (What causes these differences?)
To identify the underlying causes of overtime, we built an explanatory regression model at the monthly level. We examined the impact of various factors, such as:

• Absences,
• Understaffing,
• Characteristics of employee groups (role, team size, tenure...).

Phase 3 – Translating Insights into Impact
The results of the project were compiled into two documents: a final presentation for the steering committee and a detailed methodological report. The presentation provided clear and accessible answers to the central research questions. The results thus served as a well-founded framework for potentially adjusting HR policy based on objective insights.

The accompanying methodological report described the analytical methods employed, the filters used, and the interpretation of the results in detail. This also gave POAB a transparent insight into how the answers to their questions were arrived at. This approach makes it possible to continue working in a similar manner in the future.

Would you like to learn more, or does our approach sound like the perfect solution for your organization? Feel free to contact us!