Rutgers logo
School of Engineering
  • Events
  • Choice-Based Dynamic Pricing for Vacation Rentals

Choice-Based Dynamic Pricing for Vacation Rentals

Date & Time

Tuesday, March 19, 2024, 12:10 p.m.-1:30 p.m.

Category

Seminar

Location

Computing Research & Education Building (CoRE),

96 Frelinghuysen Road, Room 101, Piscataway, NJ, 08854

Contact

Hoang Pham

Information

Presented by the Department of Industrial & Systems Engineering

Head shot of Asian woman with long black hair.

Yaping Wang, PhD
Distinguished Data Scientist
Verizon

Abstract: We propose a new dynamic pricing approach for the vacation rental revenue management problem. The proposed approach is based on a conditional logistic regression that predicts the purchasing probability for rental units as a function of various factors, such as lead time, availability, property features, and market selling prices. In order to estimate the price sensitivity throughout the booking horizon, a rolling window technique is provided to smooth the impact over time and build a consistent estimation. We apply a nonlinear optimization algorithm to determine optimal prices to maximize the revenue,considering current demand, availability from both the rental company and its competitors, and the price sensitivity of the rental guest. A booking curve heuristic is used to align the booking pace with business targets and feed the adjustments back into the optimization routine. We illustrate the proposed approach by successfully applying it to the revenue management problem of Wyndham Destinations vacation rentals. Model performance is evaluated by pricing two regions within the Wyndham network for part of the 2018 vacation season, indicating revenue per unit growth of 3.5% and 5.2% (for the two regions) through model use.

Biography: Yaping Wang is a distinguished data scientist of Verizon. Yaping obtained her MS of statistics in 2011 and PhD of industrial and system engineering in 2012 from Rutgers University. Her current research interests include large-scale optimization in pricing and resource allocation and statistical learning in advertising and recommendation engine design.

Attendance is mandatory for in-person seminar students. For online and part-time students, seminar will be recorded and made available through Canvas.