Dynamic Pricing in Ride-Sharing Platforms
summary
Model couples strategic passengers and drivers, a platform price multiplier p(A) that depends on number of available drivers, and exponential service times so the region forms a Jackson network steady state; drivers enter based on expected earnings per time and passengers on valuation draws with a monotone hazard rate assumption. Main claim: in a large-N fluid limit optimal static pricing attains the same leading-order throughput as optimal threshold surge policy, so dynamic pricing only yields second-order gains, though it improves robustness under parameter uncertainty.
AI-written summary · about
chapters (21)
- 0:00Intro
- 1:04This talk: Pricing and ridesharing
- 2:16Related work
- 3:01Two types: Strategic and queueing
- 3:32Preliminaries
- 7:16Strategic modeling: Platform pricing
- 13:44Strategic model: Drivers
- 16:42Strategic model: Passengers
- 19:44Queueing model: Steady state
- 20:40Putting it together: Equilibrium
- 22:36The challenge
- 22:53Large market asymptotics
- 23:59What is static pricing?
- 25:29Static pricing: Illustration
- 26:41Static pricing: Interpretation
- 26:54What is dynamic pricing?
- 27:57Dynamic pricing: Numerical investigation
- 30:49Optimal dynamic pricing
- 34:39The value of dynamic pricing
- 35:46Robustness: Illustration
- 37:42Robustness: Dynamic pricing
related talks
01h 38mRidesharing
33mThe Rideshare Dispatch Problem (Aka Online Bipartite Matching with Supply Externalities)
53mCarpooling and the Economics of Self-Driving Cars
43mRebalancing in Bike-Sharing Systems – Incentives and Other Stories
44mComputational Complexity of Matching in Ride Sharing
01h 08mUnderstanding and Shaping Urban Mobility Using Dollars and Data I
52mOptimal Growth in Two-Sided Markets
44mFacilitating the Search for Partners on Matching Platforms