When machines decide the fate of employees

At Amazon, machines are often the boss.
At Amazon, machines are often the boss.PHOTO: REUTERS

(BLOOMBERG) - It is not exactly the Terminator from the sci-fi movie but in today's digital age, you can get terminated by machines, at least when it comes to your job.

Just ask Mr Stephen Normandin who was a contract driver for Amazon.com for four years.

Recently he received an automated e-mail. The algorithms tracking him had decided he was not doing his job properly and the 63-year-old US Army veteran was then fired by a machine.

He said he was punished for things that prevented him from completing his deliveries, such as locked apartment complexes.

"I'm an old-school kind of guy, and I give every job 110 per cent," he said. "This really upset me because we're talking about my reputation. They say I didn't do the job when I know damn well I did."

His experience is a twist on the decades-old prediction that robots will replace workers. At Amazon, machines are often the boss - hiring, rating and firing millions of people with little or no human supervision. Amazon became the world's largest online retailer in part by outsourcing its sprawling operations to algorithms - sets of computer instructions designed to solve specific problems.

For years, the company has used algorithms to manage the millions of third-party merchants on its online marketplace, drawing complaints that sellers have been booted off after being falsely accused of selling counterfeit goods and jacking up prices.

Increasingly, the company is also ceding its human resources operation to machines, using software not only to manage workers in its warehouses but to oversee contract drivers, independent delivery companies and even the performance of its office workers.

People familiar with the strategy say then CEO Jeff Bezos believes machines make decisions more quickly and accurately than people, reducing costs and giving Amazon a competitive advantage.

Amazon started its gig-style Flex delivery service in 2015, and the army of contract drivers quickly became a critical part of the firm's delivery machine. Typically, Flex drivers handle packages that have not been loaded on an Amazon van before the driver leaves. Rather than making the customer wait, Flex drivers ensure the packages are delivered the same day. They also handle a large number of same-day grocery deliveries from Amazon's Whole Foods Market chain. Flex drivers helped keep Amazon humming during the coronavirus pandemic and were only too happy to earn about US$25 (S$34) an hour shuttling packages after their Uber and Lyft gigs dried up.

But the moment they sign on, Flex drivers discover algorithms are monitoring their every move. Did they get to the delivery station when they said they would?

Amazon algorithms scan the gusher of incoming data for performance patterns and decide which drivers get more routes and which are deactivated. Human feedback is rare. Drivers occasionally get automated e-mails, but are mostly left to obsess about their ratings, which include four categories: Fantastic, Great, Fair or At Risk.

Bloomberg interviewed 15 Flex drivers, including four who say they were wrongly let go, as well as former Amazon managers who say the largely automated system is insufficiently attuned to the real-world challenges drivers face.

Amazon spokesman Kate Kudrna called drivers' claims of poor treatment and unfair termination anecdotal and said they do not represent the experience of the vast majority of Flex drivers.

"We have invested heavily in technology and resources to provide drivers visibility into their standing and eligibility to continue delivering, and investigate all driver appeals," she said.

When Mr Ryan Cope was deactivated in 2019, he did not argue or consider paying for arbitration. By then, he had already decided there was no way he could meet the algorithms' demands. Driving kilometres along winding dirt roads outside Denver in the snow, he often shook his head in disbelief that Amazon expected the customer to get the package within two hours. "Whenever there's an issue, there's no support," said Mr Cope, who is 29. "It's you against the machine, so you don't even try."

Amazon has automated its human resources operation more than most companies. But the use of algorithms to make decisions affecting people's lives is increasingly common.

Machines can approve loan applications, and even decide if someone deserves parole or should stay behind bars. Computer science experts have called for regulations forcing companies to be transparent about how algorithms affect people, giving them the information they need to call out and correct mistakes.

The Flex algorithms began as blunt instruments and were refined over time. Earlier, designers set too tight a time period for drivers to get to the delivery station. They had failed to factor in human nature. Drivers eager for work would promise to arrive by a certain time when they were too far away to make it. The flaw set good drivers up to fail, a source said, and was fixed only after a widespread plunge in ratings.

The system uses GPS to decide how long it should take to reach a specific address but fails to account for the fact that navigating a rural road in the snow takes a lot longer than traversing a suburban street on a sunny day.

The system worked fine for Mr Normandin for years. An Arizona native who previously delivered pizzas at night and newspapers in the morning, he knew all the short cuts and traffic choke points. He also drove for Uber and Lyft, but took on more Flex work during the pandemic when demand for rides fell and it became riskier ferrying passengers than carting packages.

He enjoyed stellar ratings and was even asked if he would like to train other drivers. He had a well-honed system: sorting packages before leaving the station, putting his first deliveries in the front seat, the next several packages in the rear and tucking the last batch deep in the back of his 2002 Toyota Corolla.

He liked gig work because he could work a few hours at a time.

Then, from last August, he had a string of setbacks he said were beyond his control. Amazon assigned him some pre-dawn deliveries at apartment complexes when their gates were still locked, a common complaint among Flex drivers.

The algorithm instructs drivers in such cases to deliver packages to the main office, but that was not open either. Mr Normandin called the customer as instructed - a long shot because most people do not answer calls from unfamiliar numbers, especially early in the morning. He called driver support, which could not get through to the customer either. Meanwhile, the clock was ticking, and the algorithm was taking note.

"There are a lot of things the algorithms don't take into consideration and the right hand doesn't know what the left hand is doing," he said.