Amazon’s Holiday Shopping List: 250,000 Workers!

Amazon just announced that they will hire 250,000 people for the holiday season. This will be a combination of full-time (I guess very few), seasonal workers, and part-time workers.

So many questions…

First… workers? Really? The Middle Ages are calling. I thought we were replacing people not hiring more... 

Second… does 250K seem like a lot or a little? Macy’s is hiring 38K seasonal workers, so maybe 250K for Amazon is not that many.

Third… can Amazon really find all these people? Are there really 250K people out there waiting for Amazon to call them… for a seasonal job?

TL;DR: Yes. Yes. And probably not. 

For the rest, let’s delve deeper: 

Workers?

I’ve written about this in the past, but it’s always worth revisiting. 

Amazon’s warehouse operations still highly depend on people, and this dependency, as well as its extent, are driven by the following factors:

1. Amazon’s business is highly seasonal.

2. Their operations support an enormous variety. 

3. Amazon’s network is highly distributed. 

Let’s look at each one in detail. 

Amazon’s sales are highly seasonal. The following figure doesn’t show the extent to which the holiday season demand is higher than the rest of the year, but it gives us an indication:

The increase in Q4 is evident, but what isn’t as clear is that the higher demand is mainly during the period between Thanksgiving and the New Year. In general, in the retail sector, about 20% of sales are generated during this period. This represents about a 3X increase in sales compared to the rest of the year. From discussions I’ve had with people at Amazon, demand during the holiday season is 6X compared to an average month in the rest of the year. Amazon tries to space things out by introducing Prime Days, but the reality is that the holiday season is significantly higher than the rest of the year. 

The two options to deal with this surge of demand are either to have a fully automated warehouse that works for most of the year with low utilization, or to add capacity as needed. Adding robotic capacity is not very feasible, so the best and most economical solution (at this point) is to hire people. People can be added with relatively little training if the processes are set up and can then be let go when they’re no longer needed. 

Amazon has an enormous variety of products. Most of these employees will be hired to work at warehouses for picking and packing orders. I’ve written in the past how this is the area where robotics has made the slowest inroads, as the reality is that picking requires qualities which, at this point, robots don’t have at the speed necessary for this type of operation. For example, we need a robot that can pick a bottle of water, a marker, and a football without breaking or spilling or dropping anything. Given the slow evolution of the dexterity of robots, this is not yet feasible at the right cost and speed, and accuracy. 

The Sparrow, depicted above, is supposed to pick and pack items: 

“... Sparrow’s robot arm dove into a bin full of random merchandise and plucked out specific items using a ‘hand’ made of small suction cups. It could identify and select merchandise buried below other items, adjusting its grip to handle different objects before depositing them in the appropriate sorting bin. Amazon says Sparrow can identify about 65% of the company’s product inventory and can tell if an item is damaged and discard it. It gets better as it learns.”

While 65% is nice, during the holiday season, 65% is not enough. 

At this point, the main advantage of people is that they can do the work for 100% of the products. Why? Because they’re flexible. 

So the first two factors (seasonality and variety) point to the simple reason why people are needed. Amazon’s operating system is built on flexibility and people offer flexibility both in terms of the ability to scale the size of operations up or down, and the ability to increase the scope of operations. 

The third factor will explain why Amazon is hiring so many people. 

250,000 Workers?

Amazon is highly distributed (close to customers). As discussed before, for many years, Amazon’s playbook was built on the ability to pool resources and enjoy economies of scale in fulfillment. Over the last 10 years, Amazon has invested in building a faster, more convenient supply chain, by being more distributed, and closer to consumers. The downside is that there aren’t significant economies of scale in such systems. What do I mean by that? A few years ago, Amazon could double the number of orders without doubling the number of employees. This is no longer the case. 

And this brings me back to the article from a few weeks ago, where we discussed Amazon experiencing diseconomies of scale, and the network’s high distribution is part of the problem. Amazon’s need for so many people is because as the system becomes more distributed, its worker efficiency deteriorates. It’s evident in the following graph how the revenue per employee has been decreasing. 

So several years ago, Amazon could add a certain number of employees for a given increase in sales. Now, for the same increase, they need to add more people since sales won’t come from one but from multiple locations, each of which is less efficient than they could be. 

So, part of the need for the additional workforce is driven by the increased inefficiency in Amazon’s operations, and part of it is driven by pure growth. And indeed, we can see the increase in the number of employees Amazon has added over the years.

It’s hard to estimate the exact number, but it’s clear that Amazon recently fired 27K employees, but hired 150K workers in 2022. With a 10% increase in sales YoY, you would expect a smaller hiring spree. But no... 

It’s also clear that Amazon is not the only firm to hire so many employees. Macy’s currently has 95K employees. If they plan to hire another 38K for the holidays, with sales that don’t outpace Amazon (judging from 2022 where Amazon and Walmart led the retail field in terms of holiday sales growth), Amazon would need to hire more than 250K employees, given its current 1.5Μ employees. 

Based on the data, it’s difficult to determine whether Amazon is conservative or bold in hiring people, but 250K seems to indicate both high expectations for the holiday, as well as an inherent inefficiency in operations. 

Can Amazon Find So Many People?

So the final question is whether Amazon can actually find 250,000 employees in such a short amount of time, and for such a short amount of time. As I write these sentences, the unemployment rate in the U.S. is 3.8%, with 1.8M people filing unemployment claims. Not all of them can work in warehouses, and not all of them want to work in warehouses. In fact, some of them probably have worked for Amazon and don’t want to do it again. 

In fact, a leaked Amazon memo from a year ago pointed out that:

“‘If we continue business as usual, Amazon will deplete the available labor supply in the US network by 2024,’ the research, which hasn’t previously been reported, says.”

Amazon confronts a potential dilemma that could significantly impact its operations. Leaked internal research from 2021, as reviewed by Recode, suggests the company might deplete its hiring pool for U.S. warehouse roles by 2024. Should this occur, Amazon’s service quality, growth strategies, and leading position in the e-commerce market could be compromised.

Amazon is considering two primary solutions to mitigate this labor shortage, which include raising wages and advancing warehouse automation. These tactics fall under the umbrella of six “levers” the company could utilize to defer the impending labor crisis. However, internal experts believe that only comprehensive changes in Amazon’s business approach and employee management can bring about a significant delay in this projected timeline.

Highlighting the specificity of this predicament, Amazon’s research underscored the immediate threat in certain regions. According to their analysis, the labor pool in the Phoenix, Arizona metro area would be exhausted by the end of 2021. Similarly, California’s Inland Empire region, proximate to Los Angeles, by the end of 2022. 

These insights grant a window into the challenges Amazon faces, revealing the contrast between its efficient online shopping facade and the intricacies of managing a vast workforce. The research underscores that Amazon’s triumph and investor appeal largely depend on its strong 1M+ workforce which operates ceaselessly.

This situation also stands as a warning for other companies looking to adopt the ‘Amazon Way’—an approach prioritizing worker productivity and witnessing high staff turnover. Historically, Amazon viewed this turnover as non-problematic or even favorable. 

We already discussed in this newsletter how Jeff Bezos, Amazon’s founder, perceived the warehouse staff as essential but substitutable, concerned that long-tenured workers might become apathetic or discontented, as highlighted by the New York Times. However, this approach appears unsustainable in the long run, especially if Amazon aspires to be the “Earth’s best employer,” a vision vocalized by Bezos in 2021. But maybe that was a different era. 

In an era where robots are knocking on our doors, Amazon is ringing the bell of human flexibility – but one wonders for how long before automation answers the call.