Demand Selection and Resource Assignment in Workforce Training Industry

The workplace training industry has grown significantly in recent years, driven by firms’ needs to keep their employees’ skills up-to-par with the rapid changes in business trends.

Also critical in today’s highly competitive and dynamic environment is matching demand with available capacity. Training firms must offer a wide variety of products and services while simultaneously ensuring high service levels and good resource utilization.

They have to decide what types of training programs to offer, how to price these programs, and the size and mix of their flexible instructional staff. Along with these choices, firms must determine whether to accept a client’s request and which instructor to assign to it.

Firms’ instructors operate at various levels of expertise. Some can teach only one or two courses, while others are able to conduct all or most of the programs. Companies must find the best way to assign their instructors to current requests without jeopardizing the chance to assign them to a future request.

Recent research from a professor at the Penn State Smeal College of Business studies how firms should dynamically respond to incoming requests for training on a certain date in order to maximize expected profits.

Susan Xu, professor of management science and supply chain management, and coauthors Yalcin Akcay of Koç University and Anant Balakrishnan of the University of Texas-Austin, find that the decision to accept or decline a client’s request depends on the number and flexibility of available instructors, the probability of different job types, and the volume and mix of future training programs.

“By accepting the current request, the firm foregoes the opportunity to use the resource for a more profitable future request. On the other hand, if a firm declines the current request, it faces the risk of having to later assign the resource to a less profitable program or leaving the resource idle on the training date,” they write.

The researchers developed a model that dynamically decides whether or not to accept an incoming job and which resources to assign to the job in order to maximize total revenue.

Of the three policies the researchers propose, they find that the Bottleneck Capacity Reservation (BCR) policy generates near-optimal solutions. It considers both the uncertainty of future requests and shifting bottlenecks, which is important for the resource allocation decisions.

The bottleneck occurs when there are too many program requests and not enough instructors to fulfill them. Firms should try to avoid these bottlenecks and ensure that their supply of resources keeps up with the demand for programs.

The researchers results show that the average profit from firms who employ the first come, first served rule is over 17 percent lower than that from the BCR policy.

“Less-flexible (e.g. specialized) resources should be used before more-flexible resources are used,” they write. “Flexible resources should be preserved for future use when capacity is tight.”

Xu and coauthors acknowledge the fact that companies are very eager to grab revenue by accepting the first requests they receive. There’s a trade-off between accepting a cheaper request up front and waiting for a more valuable request in the future.

“It’s a very realistic problem encountered in many operations management applications,” says Xu. “It’s a classical problem but because the resource structure is very complex, there is no easy solution.”

Their study, “Dynamic Assignment of Flexible Service Resources,” was published in the May–June 2010 issue of Production and Operations Management. It also received the Wickham Skinner Award for Best Published Paper.