In the world of last mile deliveries, Amazon is seen by many as an innovator. Its booming delivery network of contracted Amazon DSPs, and use of new tech in delivery systems, has set the pace for the industry, leading others to adopt parts of its practices. But some of the innovations have been harmful and not worth replicating.
Most recently, Bloomberg detailed how Amazon Flex drivers were being fired by an algorithm that determined if their performance was not up to par. In many cases, these firings occurred after events beyond the driver’s control: they were sent to deliver packages to an apartment building with locked gates at dawn, they got a flat tire, they couldn’t open a delivery locker. Even when these drivers immediately reported these issues, their ratings suffered, eventually leading to their termination.
Human beings have seemingly been removed from this management process by design and with intention. Jeff Bezos, Amazon executive chairman, is said to believe that machines make more accurate and efficient decisions than humans. Amazon has made a calculation that would be risky to any smaller business: trust the algorithms, absorb blowback related to the bot’s mistakes, replace fired drivers, and pay people to investigate firings reported as wrong. This strategy, they believed, would save them money in the long run.
Finances aside, this decision has real-world effects on drivers, sometimes putting them in dire financial situations. Amazon may be shifting HR management activities to machines, software, and bots, but we recommend proceeding with caution when considering doing the same. Some potential issues include:
- Lawsuits from employees, Department of Labor issues: Managing by bot and algorithm can lead to potentially costly errors that human judgment would not make. Amazon is able to absorb the costs of wrongful termination lawsuits related to these firings, but smaller delivery businesses could not. Most DSPs are operating on razor-thin margins, meaning that they cannot afford employment liability insurance, which would protect them in the case of employment-related lawsuits. Complaints from terminated employees could also lead to Department of Labor investigations, with the potential to really upend business operations.
- Burning through drivers: Amazon has a luxury that many delivery businesses don’t - access to labor. They are able to cycle through drivers and risk firing a good driver based on a bot’s mistake. Their Flex driver program supports this, opening driving positions up to gig workers. If their supply of drivers diminishes, they may need to rethink algorithm-based HR. Many companies are experiencing a driver shortage, meaning they don’t have the workforce to churn through. Adopting a policy that fires based solely on a software’s recommendations would compound the problem.
- Inefficient routing: When it comes to deliveries, smartly assigned routes are crucial, and many companies are currently exploring ways to utilize AI-enabled delivery software for route planning and assignments. This has potential but needs to be reviewed and monitored by human experts. We have seen firsthand flaws in algorithms sending drivers to different routes every day, causing confusion and inefficiencies.
Business leaders often look to Amazon’s innovations as the next big thing, but in this case, we can learn a lesson in what not to do. Keep humans at the center of your business and do not rely too heavily on data-driven algorithms, and bots. Flaws and mistakes could end up costing you.
Risk Strategies and our dedicated team of Transportation experts offer counsel on the biggest questions in the industry. If you’re seeking the best delivery management techniques and ideal coverage for your delivery business, reach out below.
Want to learn more?
Find Brian Jungeberg on LinkedIn, here. Find Bryan Ice on LinkedIn, here.
Connect with the Risk Strategies Transportation team at transportation@risk‐strategies.com.
Email us directly at bjungeberg@risk‐strategies.com and bice@risk‐strategies.com.
