Research
Seminars
Filter results | Clear Filters
9:00 AM - 10:00 AM
Laurens Debo and Robert Shumsky, Jeunghyun Kim
Wait Time Announcements to Increase Customer Satisfaction
-
12:00 PM - 1:30 PM
Alexandre Jacquillat
Branch-and-price for prescriptive contagion analytics
-
12:00 PM - 1:30 PM
Shi Chen
Operations Management in Cloud Computing: Pricing and Capacity Planning
-
12:15 PM - 1:30 PM
Swanand Kulkarni
TBD
-
12:15 PM - 1:30 PM
Clare Snyder
TBD
-
12:15 PM - 1:30 PM
Tauhid Zaman
TBD
-
12:15 PM - 1:30 PM
Manish Raghavan
TBD
-
Wait Time Announcements to Increase Customer Satisfaction
Speaker:
Laurens Debo and Robert Shumsky,
Jeunghyun Kim
Time: 9:00 AM - 10:00 AM
Location:
Title. Wait Time Announcements to Increase Customer Satisfaction Authors. Jeunghyun Kim (Korea University Business School), Laurens Debo (Tuck), and Robert Shumsky (Tuck) Abstract. We develop a model to optimize the provision of queue-position information to waiting customers. A customer’s emotional response, and satisfaction with the wait, may be driven by loss aversion in expected waiting time and/or by disutility due to wait-time uncertainty (we call this latter effect ‘uncertainty awareness’). We first consider three announcement strategies: no-information where a customer does not receive any updates until service begins, arrival-only where a customer is informed of the initial queue position, and full-information where a customer is informed of the initial position as well as all subsequent changes in position. Using a diffusion approximation of an M/G/1 queue, we find that for loss averse customers, arrival-only is optimal, while for uncertainty aware customers full-information is optimal. For customers affected by both loss aversion and uncertainty, the arrival-only strategy becomes more appealing as the mean service time decreases or service time variance increases. We also examine partial information, where customers are informed of their arrival to some queue positions but not all. For the diffusion approximation we find that announcements that are evenly-spaced throughout the progression through the queue is optimal. In addition, uncertainty-aware disutility is convex in the number of announcements, indicating that ‘a little information can go a long way.’Branch-and-price for prescriptive contagion analytics
Speaker: Alexandre Jacquillat
Time: 12:00 PM - 1:30 PM
Location:
Operations Management in Cloud Computing: Pricing and Capacity Planning
Speaker: Shi Chen
Time: 12:00 PM - 1:30 PM
Location:
TBD
Speaker: Swanand Kulkarni
Time: 12:15 PM - 1:30 PM
Location:
TBD
Speaker: Clare Snyder
Time: 12:15 PM - 1:30 PM
Location:
TBD
Speaker: Tauhid Zaman
Time: 12:15 PM - 1:30 PM
Location:
TBD
Speaker: Manish Raghavan
Time: 12:15 PM - 1:30 PM
Location: