Problem:
A large convenience store chain wanted to introduce fresh food production, new layouts, self-checkout, and delivery across thousands of stores. However, small changes in one location often disrupted convenience store operations in another.
Solution:
Mosimtec built a realistic virtual store using AnyLogic to test store layout design, staffing, and customer behavior before making real-world changes.
Results:
- Showed when fresh food boosts profit and when it strains staff.
- Clarified where self-checkout helps and where it falls short.
- Helped balance labor, delivery demand, and in-store flow.
- Gave franchisees and leaders a shared, data-backed way to make decisions.
- Reduced risk before scaling changes across the entire network.
Introduction: testing changes at scale
Mosimtec, a consulting firm focused exclusively on modeling and simulation, partnered with one of the largest convenience store chains in North America on a major transformation initiative.
Client’s convenience store chain operations overview
The client with over 1,000 stores was looking to rethink convenience store operations at a day-to-day level. To move this forward, a major North American retailer needed a way to test new ideas, including changes to store layout design, before scaling them across thousands of locations.
Challenge: rethinking convenience store operations
Approximately 80% of the client’s convenience stores sell fuel, and for those locations, it accounts for roughly 60–70% of store revenue. While fuel drives volume, it is a low-margin business. The real opportunity lies inside the store, where higher-margin products can increase profitability and reshape convenience store operations.
Customer behavior is also changing. Traditional convenience items are losing appeal, especially among younger customers. Fresh food offerings provide a higher-margin alternative and a reason for customers to enter the store rather than remain at the fuel pump.
But turning that idea into reality was complex.
The client operated thousands of stores with different layouts, volumes, and regional preferences. Decisions around store layout design that worked in one location often failed to translate to another.
Fresh food production changed how stores ran. New equipment improved quality but slowed preparation. Cooking, keeping prepared food warm and available for sale, and managing waste became part of everyday operations. At the same time, a new loyalty app and delivery partnerships introduced demand patterns that were harder to predict.
Several challenges surfaced quickly:
- Higher labor requirements.
- Demand spikes from delivery orders.
- Mixed results from self-checkouts.
Many stores were independently operated by franchisees running lean operations. The client needed a reliable way to test whether these changes would pay off before scaling them across the network and further altering convenience store operations.
Read also: How simulation helped a large retailer optimize checkout processes and evaluate self-checkout performance before implementing changes across multiple stores.
Solution: testing new store layout design at scale
Mosimtec built a virtual convenience store in AnyLogic to model customer and employee movement, congestion, and queues across different store layouts. The goal was not to model an idealized store but to capture real behavior, real constraints, and real trade-offs that shape convenience store operations.
A flexible, configurable store model
The model was built to work across a large and diverse store network. Instead of locking assumptions into the logic, most store characteristics were controlled through data inputs.
This included:
- Store layouts and footprints.
- Product zones and merchandising placement.
- Menu options and food equipment.
This flexibility allowed teams to test different approaches to store layout design side by side, even when locations varied significantly in size, layout, and customer mix.
Realistic customer behavior
Rather than focusing only on foot traffic, the model captured how customers actually used the store:
- When customers arrived throughout the day.
- What they purchased.
- How long they spent in different areas.
Once fresh food and delivery orders were introduced, these details became especially important, as they directly influenced queues, congestion, and staff workload within daily convenience store operations.
Workforce at the center of operations
Labor was treated as a shared resource rather than a fixed schedule. Employees moved between tasks the same way they do in real stores.
The simulation accounted for:
- Task switching and interruptions.
- Food preparation and holding.
- Cleaning, deliveries, and other non-checkout tasks.
Read also: See how Domino’s used simulation to test store layouts and staffing, helping teams understand what actually works before making changes in real stores.
Visualizing movement and congestion
The model tracked how customers and employees moved through the store. Congestion was clearly evident around coffee stations, food areas, and checkout zones, directly tying movement patterns to the store layout design.
Teams tested layout changes in the simulation before committing to physical renovations.
This gave the client a practical way to evaluate decisions and understand trade-offs before introducing changes across the store network.
Results: impact on convenience store operations
The simulation showed that the fresh food initiative can succeed, but only with the right operating model in place. Stores running with minimal staffing struggled to support food preparation and delivery without disruption. When staffing levels increased, higher-margin fresh food offset the added labor costs and strengthened overall convenience store operations.
The model also reset expectations around new initiatives:
- Self-checkout performed well during certain periods but delivered limited value during peak evening hours due to age-restricted sales.
- Delivery demand remained manageable when stores were staffed and equipped to handle clusters of orders.
Beyond operational insights, the simulation helped align stakeholders. Franchisees recognized their day-to-day challenges reflected in both staffing assumptions and store layout design. This shifted discussions away from opinions and toward evidence.
Instead of debating assumptions, teams tested scenarios and evaluated trade-offs directly. For a transformation affecting thousands of stores, this approach reduced risk and supported more confident, informed decisions.
The case study was presented by Nelson Alfaro Rivas from Mosimtec at the AnyLogic Conference 2025.
The slides are available as a PDF.