Electronic Commerce System

Client: National Retail Federation Conference

Project Statement

This model was presented at a conference as a demonstration of how SIMPROCESS can be applied to eComm implementation. It represents a fictitious company that receives orders for three products: A, B, and C. In the model, each order will be processed twice, once using the “Brick and Mortar” approach, and the other time incorporating real-time access to the company’s inventory levels, a “Click and Order” approach.

The focus of this model was to show how the ability to have real-time access to existing inventory provides a better estimate of the delivery date. If an item is in stock, it will be delivered faster, and more accurate information of inventory levels will lead to higher accuracy of the predicted delivery time.

The model does not have much depth, but it shows the “As-Is” process (Brick and Mortar Process) and the “To-Be” approach (Click and Order Process) running at the same time with the order numbers updated on the screen. Building to a ‘real’ model from a demo version would allow the user to build the business case and determine the number of operators needed, as well as quantify savings.

CACI Services Involvement:

SIMPROCESS was used to build the simulation model


Model was for demonstration purposes only.


This demo model was run for approximately two and a half years (simulation time). In that time, 67,737 orders were placed for either product A, B, or C. The two different processes produced dramatically different results.

The Brick and Mortar Process predicted 75% of the items ordered were in stock, and therefore would be delivered by UPS within two business days. The other 25% were not in stock, so they would be delivered within five business days. However, when the simulation was run, the company was not that accurate with its results. Almost 15% of their deliveries were later than promised. The company predicted 75% of the items would be delivered within two days, but only 70% actually were. Likewise, 25% of the orders were 5-day deliveries, but only 15% arrived in time.

The Click and Order Process produced much more favorable results. The company predicted 80% of the orders were in stock and would be delivered within two business days. The simulation run showed close to 78% of those orders arrived in the time promised. In addition, the company promised the other 20% of the orders would be delivered within five business days, and close to 17% were delivered in this time frame. In the Click and Order Process, only 5% of all orders arrived later than expected.

The strength of simulation modeling comes from comparing the two processes. The Brick and Mortar Process produced three times as many late deliveries as the Click and Order Process, a total of 6,741 late deliveries. It is safe to conclude that having real-time access to a company’s inventory levels produces more accurate delivery predictions. In addition, SIMPROCESS would allow a company the ability to quickly configure the current demo model to more closely resemble their actual process, and look at different configurations of resources, product flow, etc.

It is important to remember in this case that this was a demonstration model only, and it would require much more input to accurately represent an actual business process.