The controversy surrounding Uber has called the usefulness of customer reviews into question. Is it time to modify the system?

Written by Michael Thomsen (@mike_thomsen)


"They're running a sweatshop with an app," Raj Alazzeh told Liz Gannes of AllThingsD. Alazzeh had worked with the newly started Uber service, a cell phone app that lets users order a car without having to dial any specific service or talk to a dispatcher, and had come to its San Francisco offices to protest its policies. Uber doesn't technically employ any of its drivers, but instead funnels users to drivers it registers through a database. The app allows for user ratings of individual drivers and a few bad ratings can be enough to delist drivers or lower their pay, according to Allazeh. It also makes drivers compete with noncommercial drivers who don't carry commercial liability insurance, something the California Public Utilities Commission (CPUC) eventually approved but longtime cabbies aren't happy about. Alazzeh claimed 500 drivers had been cut from the service in February because of customer reviews.

Since its launch in 2012, the app has encountered fines and legal struggles with its disruptive aspirations for the heavily and variously regulated taxi services around the country. The company was fined $20,000 by CPUC for operating without a license and is contesting lawsuits in San Francisco and Chicago. In New York, the company faced several conflicts with policy, such as the prohibition on taxi drivers using electronic devices while driving as well as unjustifiably refusing to pick up fares. It faced local criticism for the practice of "surge pricing" after Hurricane Sandy, doubling fares after the natural disaster at a time when other companies were offering freebies and waving fees. 


Apps like Uber use technology to normalize mistrust and create a venue where it can be used as social currency.


Uber's CEO Travis Kalanick said the claim of 500 delisted drivers was "more than an order of magnitude off" but said the company could not continue to list drivers "who don't know the city well" or "receive consistently negative feedback from riders." Because Uber functions as a connectivity tool rather than an actual transportation service, customer ratings are the most immediate way they have of evaluating drivers, and cutting pay or delisting drivers their primary method of quality assurance.

The standoff is a model for how technologies that increase systemic efficiency don't necessarily improve the system as a whole. Apps like Uber identify one narrow element of a healthy urban environment and create a hyper-efficient channel for access to it while off-loading driver screening, car maintenance, ongoing training, and employee benefits to the workers or the secondary services that employ them. It diminishes the complexities of evaluating and supporting a person's work contributions to a comment box that encourages its customers to view their experience in the most depersonalized way possible.


Relying on customer reviews is a dubious practice, from Yelp's infamous practice of targeting poorly reviewed restaurants with offers of special advertising packages to improve their image, to the enlistment of hired reviewers to write falsely enthusiastic reviews. Yet customer reviews are essential for technologies like Uber and how they project an aura of transparency and fairness, which always seems to come at the expense of the workers expected to live up to these standards with only the disembodied fragments of computer text to go by. For customers, the review box becomes a way of avoiding a potentially awkward social encounter by telling a person face-to-face what they'd needed or wanted, and instead encourages complaining ex post facto while disconnecting from any responsibility for the impact of those complaints on another person's living conditions.

These modifications of social connection create an environment of increasing antagonism between individuals, encouraging a factional separation between customer and service provider, not because either party wants to view the other in that way, but because the technical intermediary is only capable of depersonalized characterizations. Apps like Uber use technology to normalize mistrust and create a venue where it can be used as social currency. There's no technical reason Uber couldn't instead ask its users what their favorite thing about their driver was, or have them describe an interesting memory they had from their drive home.

There is no efficient purpose to creating a storehouse of this kind of information, but it would reflect a fundamentally different set of social values that say the worker is more important than the work. But technology only gives us what we put into it, and so we get a paradoxical rise in efficiency and dissatisfaction with each new disruption. And the intermediary becomes invisible so instead of blaming it for its uselessly narrow standards and insufficient involvement with the workers whose labor they are selling, it encourages worker and buyer to be skeptical of one another while making it impossible for either to imagine connecting with the other without an app.