Odds are you’ve heard that Amazon plans to make a huge investment in retraining its existing workforce, partly due to the displacement of employees by emerging automation and A.I., and partly due to scarcity of talent in key job families.

I want to take a look at the Amazon re-skilling investment with a critical eye, but first here’s a primer of what Amazon has planned for the uninitiated:

"Amazon (AMZN) today pledged to upskill 100,000 of its employees across the United States, dedicating over $700 million to provide people across its corporate offices, tech hubs, fulfillment centers, retail stores, and transportation network with access to training programs that will help them move into more highly skilled roles within or outside of Amazon.

Amazon’s Upskilling 2025 pledge invests in a range of new upskilling programs to serve employees from all backgrounds and Amazon locations. Programs include Amazon Technical Academy, which equips non-technical Amazon employees with the essential skills to transition into, and thrive in, software engineering careers; Associate2Tech, which trains fulfillment center associates to move into technical roles regardless of their previous IT experience; Machine Learning University, offering employees with technical backgrounds the opportunity to access machine learning skills via an on-site training program; Amazon Career Choice, a pre-paid tuition program designed to train fulfillment center associates in high-demand occupations of their choice; Amazon Apprenticeship, a Department of Labor certified program that offers paid intensive classroom training and on-the-job apprenticeships with Amazon; and AWS Training and Certification, which provide employees with courses to build practical AWS Cloud knowledge that is essential to operating in a technical field."

700M is a lot of money. Let’s do some simple math and then start evaluating how to the investment could intensify if it wasn’t spread evenly (which is never is):

--First the simple match.  700M across 100,000 impacted employees equals a base investment in retraining/upskilling of $7,000 per employee. Compare that to the average annual per employee investment in Learning and Development cited by Bersin ($1,200), and the investment seems solid above and beyond what Amazon already does.

--Now imagine a world where the investment isn’t spread out equally across all employees.  Since the Amazon upskilling initiative will have a voluntary vibe to it (similar to AT&T’s upskilling efforts require the employee to proactively opt in and spend their own time preparing their skills for the future), it’s not hard to imagine the opt in rate won’t approach anywhere near 100%. 

--Spread the 700M investment over 50% of the employees, and you’ve got an investment of $14,000 per employee.

--Spread the 700M investment over 30% of the targeted employees, and you’ve got an investment of over $23,000 per employee.

The devil, as it always is, is in the details.  It's a cool program. Will Amazon spend the same total amount of money if just 30% of the impacted employees opt in to the program? The presence of pre-paid tuition and certification programs suggests no.

The voluntary, opt-in nature of the Amazon Upskilling 2025 program is necessary. After all, employees impacted by A.I. and automation have to WANT to improve their long term career prospects. That's why so much of this program will have to be completed after work hours.

That's going to sound like a second job (unpaid as well) to a lot of employees. That means Amazon likely won't spend as much as projected.

If you were in Vegas, you'd take the "under" related to the bet of whether Amazon will spend more or less than 700M by the year 2025 on this program.