Abstract: Agent-based modelling and simulation offers a new and exciting way of understanding the world of work. In this paper we describe the development of an agent-based simulation model, designed to help to understand the relationship between human resource management practices and retail productivity. We report on the current development of our simulation model which includes new features concerning the evolution of customers over time. To test some of these features we have conducted a series of experiments dealing with customer pool sizes, standard and noise reduction modes, and the spread of the word of mouth. Our multi-disciplinary research team draws upon expertise from work psychologists and computer scientists. Despite the fact we are working within a relatively novel and complex domain, it is clear that intelligent agents offer potential for fostering sustainable organisational capabilities in the future.
Introduction
The UK retail sector faces a persistent productivity gap compared to international competitors. While traditional Operations Research (OR) focuses on logistics and inventory optimization, it frequently overlooks the "human element"—the complex ways in which management practices like training and empowerment influence customer behavior. To address this, researchers have developed ManPraSim, an agent-based simulation (ABS) designed to bridge the gap between human resource management and retail productivity.
The ManPraSim Framework
Unlike standard modeling tools that provide static snapshots, ManPraSim utilizes Agent-Based Simulation to capture the proactive and stochastic nature of human decision-making. Implemented via the AnyLogic platform, the model focuses on the autonomous interactions of individual agents rather than a simple flow of entities through a process.
As shown in the architecture above, the model integrates various agent types—Customers, Sales Staff, and Managers—with global parameters to generate emergent macro-level behaviors and identify operational bottlenecks.
Modeling Customer Logic and Satisfaction
A core innovation of this research is the use of State Charts to define agent logic. These charts allow customers to transition between states such as "Browsing," "Seeking Help," and "Queuing." To quantify the customer experience, the authors introduced a Service Level Index, which assigns satisfaction weights to specific transitions.
In this logic flow, positive experiences (like receiving help immediately) grant +2 satisfaction, while negative outcomes (like leaving a queue in frustration) result in a -2 penalty. This allows the simulation to track cumulative satisfaction over a customer's "lifetime" within the model, reflecting how past experiences influence future return rates.
Empirical Grounding and Methodology
The model is calibrated using empirical data from four departments across two branches of a major UK retailer, specifically focusing on Audio & Television (A&TV) and WomensWear (WW). These environments provide contrasting data points: A&TV involves high-cost items and long service times, while WW is characterized by high volume and "shopping enthusiast" behavior.
Key technical features include:
- Noise Reduction Mode: A custom execution mode using constant arrival rates to isolate the impact of specific variables from daily retail fluctuations.
- Word-of-Mouth (WOM) Algorithm: A mathematical model that calculates the adoption rate of new customers based on the ratio of satisfied to dissatisfied shoppers from previous cycles.
Key Insights and Results
The simulation revealed several critical insights into retail dynamics:
- The Double-Edged Sword of WOM: While positive word-of-mouth increases footfall, it can decrease overall satisfaction if staffing levels remain static. Increased traffic leads to longer queues, which eventually offsets the gains from marketing.
- Staffing Bottlenecks: In the WomensWear department, "Normal Mode" (with real-world noise) showed a 25% increase in customers leaving while waiting to pay compared to "Noise Reduction" mode, highlighting how staffing inflexibility creates major bottlenecks during peak hours.
- The Sunday Bottleneck: The model revealed that the A&TV department struggled on Sundays because their "dedicated cashier" model could not handle the specific arrival patterns of that day, a failure often hidden by standard averages.
Conclusion and Future Work
While ManPraSim successfully demonstrates how micro-level interactions drive macro-level business outcomes, the authors note that current staff agents are somewhat passive. Future iterations of the model will focus on "Staff Empowerment," creating proactive agents capable of switching tasks based on real-time demand, and refining how internal stimuli and teamwork dynamics influence the overall retail ecosystem. Despite the complexity of the domain, intelligent agents offer significant potential for fostering sustainable organizational capabilities.
Summary updated