How Amazon automatically tracks and fires warehouse workers for ‘productivity’

By Oleg Mikhailenko
Published in E-Commerce Market
March 14, 2022
5 min read
How Amazon automatically tracks and fires warehouse workers for ‘productivity’

Amazon prides itself on its performance management system which claims to keep its warehouses running efficiently and its employees on task. Artificial intelligence has increasingly been utilized as a means for Amazon to track its employees and increase productivity. However, reports from employees paint this technology in a darker light. Instead of being used as a vector for efficiency, cheerfully maintaining a quick pace throughout Amazon’s fulfillment centers, workers report feeling monitored, rushed, and objectified - as if the company views them as more machine than human. From reports of automatic write-ups or terminations to rumors that Amazon employees even relieve themselves in bottles to avoid falling below mandated picking times, it’s clear that employees view the use of this personnel managing software with distaste and even resentment.

The question, therefore, is as follows: How does this software function? Does Amazon use artificial intelligence to make hiring, disciplinary, and termination decisions at all, as many workers claim? If so, is it the case that this program only makes initial decisions which are then followed up upon by a human being, or does what the computer says, go? How does Amazon use its AI on a day to day basis?

Amazon claims that their “proprietary productivity metric”, as they put it, is based on objective measures such as customer demand and warehouse location. Associates are required to be detailed and efficient in processing orders so as to make sure that everything runs smoothly. Through the use of convenient handheld scanners, Amazon claims that it gathers select amounts of intelligence to ensure that everything is running as it should. Employees, on the other hand, experience a heavy pressure to “make rate”. One employee reports that he has a scant fifteen seconds to scan items and place them into the correct carts. On his night shift, his main interaction is not with other human beings, but with machines. Indeed, Amazon is even reported to use (frequently glitchy) algorithms in regards to hiring.

Packing hundreds of boxes, day in and day out, is a strenuous task for any employee. At the quick pace Amazon demands, however, it can feel almost unbearable. Employees report to believe that they are being constantly monitored by cameras as another level of surveillance on top of their scanners. Select fulfillment centers are even rolling out wristbands that allow Amazon to track employees on a finer level than ever before.

Between August of 2017 and September of 2018, a spokesperson for Amazon reported that the company had fired 300 full-time employees at only a single facility. Assuming this rate held steady today, this would imply that Amazon is firing or laying off approximately ten percent of its employees per annum, for productivity reasons alone. However, a stark increase in terminations and leaves of absence due to the Covid-19 pandemic have also been documented, suggesting that this rate is even higher than expected. With almost one termination for every day of the year, it begs the question as to how managers find the time to process all the disciplinary action leading up to each firing.

To hear the employees tell it, the simple answer is that they don’t. Workers report that they are automatically monitored and timed during their work day and any portion of their time on the clock deemed TOT, or “time off-task”, is logged. This can, and frequently does, lead to instances wherein an employee can be automatically written up by Amazon for productivity concerns, and eventually fired without the decision ever being reviewed by a human being. This would be bad enough even if the algorithm were wholly accurate, and there was never a case where Amazon’s system was down or malfunctioning. Unfortunately, Amazon’s employee tracking software is notoriously buggy, frequently leading to mistaken disciplinary action or even firings that later prove to be unjustified. Appeals are theoretically possible, but the time they take to process through the unreliable system means weeks or even months of missed income for employees and their families.

The threat of going without an income over the matter of a few seconds weighs heavy on the average employee. In the shadow of Amazon’s use of automated software, workers have even been reported as relieving themselves in bottles and bags, which are littered around several fulfillment centers. Although Amazon largely denies these allegations, the number of employees reporting such occurrences leverages serious questions about their trustworthiness on this matter.

The allegations that Amazon is working with the Pinkerton Spy Agency are equally concerning. Pinkerton, which has a long and storied history of often violent anti-union efforts, is reported to be partnering with Amazon to monitor worker organization in its European fulfillment centers, which is, in many cases, in direct violation of EU worker rights legislation. Although no fulfillment centers in the United States are currently unionized, economic commentators have posited that it’s extremely likely Amazon is closely monitoring union activity in the US as well. Union activity which might conceivably negotiate towards the reduction of breakneck packing rates, or even demand Amazon break ties with its automation companies running the software altogether, is considered extremely undesirable by the e-commerce giant.

It’s plain to see that many employees are fed up with feeling as if their jobs are at the mercy of Amazon’s AI. Although public health experts have attributed Amazon’s high rate of injury amongst its workers (which hovers between 5.9 and 6.5 incidences per 100 workers, more than twice that of Walmart), there is at present little in the way of promising legislation aimed at regulating these unmanageable workloads. Although Former CEO Jeff Bezos has gone on record stating that Amazon needs to do better for its employees and expressing a desire for the e-commerce juggernaut to become “the best employer in the world”, due to increased demand from consumers in the wake of the Coronavirus pandemic, workloads have only increased, alongside mandated picking speeds.

The tension between high workload demands, staffing shortages due to Covid-19 and worker injuries, and the increased scrutiny on Amazon of late in the aftermath of tragedies such as the destruction of a fulfillment center in Edwardsville, Illinois by a tornado, seem to be causing more friction every day. It seems more a question of ‘when’, rather than ‘if’, an external party such as a federal oversight agency or even the United States Congress itself will step in if Amazon neglects to act. In the lead-up to the United States midterm elections, it’s highly likely that worker-friendly legislation will be introduced to secure votes from Amazon’s many employees, especially given the steadily-mounting demands during this, the third calendar year of the Coronavirus pandemic.

Although efficiency is paramount in any warehousing or fulfillment environment, Amazon’s business needs do not and should not come at the expense of the wellbeing and health of employees. Indeed, it is only through the fair treatment of workers that companies can continue to thrive in the age of the gig economy and the Great Resignation. As the reputation of automated performance management systems such as the ones Amazon uses to track its employees has been damaged by the association, it may be the case that this technology falls out of favor in the future in favor of other means of improving warehouse efficiency.


Tags

#amazon#management#AI#employees#intelligence#productivity#tracking
Previous Article
Market Integration

Oleg Mikhailenko

Co-Founder, CEO P2Pseller

Related Posts

Ebay’s seller hub: what you need to know
August 01, 2022
2 min
P2PSeller Logo
LinkedinTwitter
© 2022, Modern Trading Technologies, Inc. USA, California. All Rights Reserved.