Uber tests system that estimates how much customers are willing to pay for a ride

Uber is working on a system that can predict how much customers are willing to pay for a given ride. The software calculates the amount to be paid based on, among other things, the location of departure and arrival, and the time.

That’s what Daniel Graf, who works at Uber as the head of the Product department, said in an interview with Bloomberg. According to the CEO, a system is being tested in some cities that, among other things, uses data about the journey and the time to predict what a customer would like to pay. For example, the system thinks that customers who take a ride from a wealthy neighborhood to another rich neighborhood are willing to pay more than customers who take a ride within poorer neighborhoods.

More detail was not given about the so-called route-based pricing, but according to Graf it is a complex system that works in the background; when a customer requests a ride, he or she can immediately see how much has to be paid for it. For now, though, that’s only being done in a few US cities, and Uber hasn’t said if or when it’ll be rolled out globally.

Graf’s explanation comes after a number of Uber drivers complained about the growing discrepancy between what the customer pays and what the driver receives for the ride. According to Graf, the prediction system is the reason for this, and Uber would need the money to cover the losses.