Decision, Operations and Information Technologies
The Dynamic Price Is Right
Apr 02, 2018

The Dynamic Price Is Right

Robust Optimization Model Helps Retailers Adjust Discounts to Match Demand

Apr 02, 2018
Decision, Operations and Information Technologies
As Featured In 
Manufacturing & Service Operations Management

Retailers face a range of dilemmas when setting prices for trendy items like consumer electronics gadgets and fashion apparel. A new robust optimization model developed at the University of Maryland’s Robert H. Smith School of Business can help store managers make better decisions about when to offer discounts and how low they should go.

“Time-sensitive products have short selling seasons and long supply lead times, so retailers cannot simply wait until supplies runs low and then order more,” says Smith School professor Zhi-Long Chen, co-author of the study, featured in Manufacturing & Service Operations Management. “No inventory replenishment is possible during the compressed selling season.”

Unlike common household products such as food, beverage, personal hygiene products and home appliances, which have a long product life cycle and hence quite predictable demand, little historical data are available for the latest cameras, smartphones or designer bags. Store managers must make educated guesses instead about demand.

One tool available to maximize profit while facing demand uncertainty is dynamic pricing. “The decision maker needs to sell a given amount of inventory of each product by periodically adjusting prices over the selling season,” the co-authors write.

Additional complexities arise when consumers have substitute choices available side by side on store shelves. “Products do not exist in isolation,” Chen says. “Some products are similar in nature, so price adjustments to one product can effect demand for another.”

Existing dynamic pricing models account for this effect. But they overlook a second type of substitution that occurs over time, when demand shifts for a single product or multiple products during different periods within the selling season. 

Chen and his co-author, Ming Chen at California State University, Long Beach, account for both types of substitution. “Our model is much more practical than most existing multiproduct dynamic pricing models,” they write. “We consider both interproduct and intertemporal demand substitutions.”

The authors also account for real-world factors such as logistical constraints and consumer psychology. For example, most retailers understand that adjusting prices too often or too dramatically can backfire. “Frequent price changes and significant price differences from one period to the next may confuse the consumers who are searching for appropriate prices, not necessarily the lowest ones,” the authors write. “It has long been understood that frequent price changes can even make a retailer appear unfair or dishonest.”

No simple pricing rules work for every case, but the robust optimization model developed by the authors provides a relatively straightforward tool for finding optimal or near-optimal prices for the products over time with a reasonable amount of computation. The model also allows for a relatively wide margin of error, which results from over- or underestimating demand.

“No precise knowledge of the demand bounds is required to use our model to generate satisfactory solutions,” the authors write.

Read more: Robust Dynamic Pricing with Two Substitutable Products is featured in Manufacturing & Service Operations Management.

About the Author(s)


Dr. Chen received his PhD degree in Operations Research from Princeton University in 1997. He is currently Orkand Corporation Professor of Management Science at the Robert H. Smith School of Business. Prior to joining the Smith School in 2001, Dr. Chen worked as an assistant professor of systems engineering at University of Pennsylvania for four years. His research interests cover supply chain scheduling, production and transportation operations, dynamic pricing, and optimization. Dr. Chen has conducted several NSF funded research projects on integrated production and distribution operations, coordination of dynamic pricing and scheduling, and transportation capacity planning.

How Sellers Can Better Understand Demand
For B2B sellers, knowing how much the buyer is willing to pay is difficult. New research from Maryland Smith is helping to figure it out.
Jan 05, 2021
The Issue Companies Have With Moving Up
New research shows service providers at the lower end of the value chain mostly find that their efforts to move up yield limited success.
Dec 15, 2020
What Online Retailers Get From Recommendation Lists
New research shows with online shopping, retailers’ use of recommendation systems increases their overall sales and consumer click-throughs and conversion rates. The effect is greatest for mobile shoppers.
Dec 15, 2020
Robert H. Smith School of Business
Map of Robert H. Smith School of Business
University of Maryland
Robert H. Smith School of Business
Van Munching Hall
College Park MD 20742