Transportation & Logistics
Case StudyUber

Uber discriminates price for different customers using machine learning and a variety of sociological data

Uber has piloted personal price discrimination using machine learning and based on sociological data including neighbourhood affluence. Previously price fluctuated based on demand, mileage, and time of day.

Context

Uber is a ridesharing app that matches customers looking for a car ride with drivers nearby offering to drive them. Uber piloted "route-based pricing" in 2017 to better price discriminate based on more customer data: "In the past, your fare would be generated based on mileage, time, and geographic demand". (ZDNet)

The Project

The new pricing system estimates fares for passengers based on sociological factors in addition to time of day, starting point and end point. Machine learning can facilitate this first degree price discrimination because it "is the hardest to enforce, as it requires the seller to have perfect information on the customer... By leveraging knowledge from its in-house economists and statisticians, Uber has developed an effective algorithm on which to base fare prices". (Accenture)

Data

From Accenture: "[S]eries of aggregated data variables such as time of day, route requests [passenger location and destination] and neighbourhood wealth" along with historic data of previous passenger requests.

Results

Results undislcosed, however ZDNet reports that "Uber pockets the leftover amount between the driver''s pay and what a customer is charged", so theoretically the pricing change will increase profit margins.

Back to Case Studies
AI Daily Brief — leaders actually read it.

Free email — not hiring or booking. Optional BPAI updates for company news. Unsubscribe anytime.

Include

No spam. Unsubscribe anytime. Privacy policy.