Unionization at Amazon and the future of employee surveillance

All eyes were on Bessemer, Alabama this week, as workers at Amazon’s warehouse voted on whether or not they would unionize. Although the historic attempt was defeated, the demands made by union proponents raise broader questions around the future of workplace surveillance, both for the warehouse workers at Amazon and more broadly.

In addition to questions of pay and benefits, one of the main drivers of unionization efforts was the lack of human dignity in the workplace brought about by Amazon’s constant surveillance. Tracking productivity metrics, such as Amazon’s “time off task,” is crucial for optimizing operations and ensuring deliveries are fast and low-cost. However, by ruthlessly focusing on tracked metrics, Amazon has left a number of employees feeling overwhelming pressure and stress to reach metrics-based results.

Although Amazon struck down its current unionization attempts, this does not prevent workers from trying to unionize in the future, both at Amazon and elsewhere. Employers, as such, should be galvanized by this event to examine their own strategies for workplace data tracking and surveillance, to ensure they are maintaining human dignity while balancing for productivity.

Based on the controversy surrounding Amazon’s surveillance mechanisms, I believe there are a few key lessons learned that should be implemented at other organizations collecting surveillance data, including the following:

  1. Start with a culture of trust – The way employees react to surveillance is largely driven by their overarching views of the organization. If they see the company as a collaborative, learning-oriented organization, they will likely view the outputs of surveillance data in the same vein. However, if there is a lack of trust in the organization, employees will draw their own conclusions, fearing promotion and firing decisions will be based on metrics to which they might not have full visibility. Organization should push to build trust from the bottom-up before implementing large employee-tracking initiatives
  2. Be clear on objectives – At Amazon, there is a relentless focus on optimization, which drives the company’s rationale for monitoring. However, employees do not always have a clear understanding of the expectations with regards to performance. For example, many employees have cited constant fear of termination based on an unknown threshold of their individual “time off task” metrics. If companies are setting boundaries on unacceptable metric outputs, they should clearly inform their employees, so they can target these goals
  3. Help employees improve – Similar to a lack of clarity on objectives, Amazon employees felt additional stress after seeing individuals terminated due to their “time off task” metrics. Rather than rushing to terminate, companies should keep employees in the loop on their metrics and provide them with the tools and resources to improve. This can include training, mentorship, and feedback from top-performers
  4. Gather feedback to understand what the metrics don’t track – No one metric can provide a holistic understanding of how employees are spending their time. If a company seeks to track productivity via a simple metric, it is critical that they gather feedback to understand the full context and what the metric isn’t measuring. For example, “time off task” at Amazon includes all non-packing activities as off-task, even if they are critical responsibilities associated with the role. By gathering feedback from employees, companies can learn what else may be driving these metrics and understand them more holistically.





Keeping Performance Reviews Honest: Lessons from the Marines


Nestlé: People Analytics and Gender Equity

Student comments on Unionization at Amazon and the future of employee surveillance

  1. There are many ethical arguments to be made against Amazon’s increased surveillance practices, yet one less-discussed issue is an economic one; where, rather than sharing the economic value generated by these programs with employees, the company itself captures all of it.

    Every job has a certain gap between the optimal and actual productivity levels. After all, we are all human; and it’s only normal to have an occasional bad day or take a slightly longer break at work. With increased surveillance and rules built around them, Amazon aims to standardize employee performance by creating a risk of job termination, effectively aiming to remove the few seconds workers take to recollect their thoughts, the extra bathroom break taken on a given day, and many similar variations in worker behavior.

    The result is that workers significantly increase their mental and physical workload, experience more stress, and on average work harder than they would have otherwise. However, the credit for this increase in harder work is taken by the company itself. Employees are not getting salary increases for working harder, nor receive additional paid vacations for finishing tasks earlier than they would before these rules came into play. They are just assumed to “finally” start performing at the level that their job description asks them to perform at.

    This is very problematic and will only cause further tensions between the company and its workers – particularly in the US, where worker protection laws are less robust and the ability of workers to act collectively through unionization is highly politicized and controversial.

  2. I could not agree more. Amazon’s relentless push to optimize operations negated the human element, stressed and overwhelmed employees, and led them to unionize. Going forward, your recommendations on building trust, setting clear objectives, helping employees improve, and gathering additional data are an excellent starting point.

    One issue that I think Amazon will face at its warehouses is building trust in a high turnover environment. Amazon uses seasonal workers to augment its full-time staff during peak demand – holiday season, Prime Day, etc. Further, Amazon employs the third largest workforce in the United States, behind the U.S. government and Walmart. To build trust, Amazon will have to create scalable policies that empower managers and employees alike. It will be interesting to see how Amazon changes its policies following Bessemer, Alabama’s unionization efforts.

  3. This is a super interesting post! I’m curious how you think companies should balance transparency with the desire to continue to modify metrics as they develop. Should companies notify employees when they make changes to how data is used? I’m especially curious in the context of a company like Amazon that is under heightened levels of external scrutiny. Opening the “black box” of an algorithm risk exposing a company to criticism and liability about how they’re using their data, but I think you’re totally right that some amount of transparency is critical to building trust. I’m also struggling a bit to think about how much workers “should” be willing to give up in terms of privacy and tracking in exchange for better pay and benefits–it’s clearly the argument that prevailed in this case, but I think there are real justice concerns about how much society “should” allow and what responsibility companies have to notify and explain things to their employees.

Leave a comment