Amazon Go Is Dazzling. But How Many Jobs Will It Kill?

SEATTLE, WA - JANUARY 22: Shopper Ela Ustel walks through the Amazon Go store, on January 22, 2018 in Seattle, Washington. After more than a year in beta Amazon opened the cashier-less store to the public. (Photo by Stephen Brashear/Getty Images)
Shopper Ela Ustel walks through the Amazon Go store on Monday in Seattle. Stephen Brashear/Getty Images

Amazon Go, the Seattle supermarket built to get rid of lines inside the store, opened for business on Monday with a big line outside.

The place is by all accounts a marvel: an 1,800-square-foot convenience store that sells sandwiches, snacks, beer and more, but instead of checking out, you check in, by scanning a smartphone code at a glass-doored entry gate. When you’re done shopping, you just walk right out, and Amazon sends you a receipt. Just as it does online, Amazon records your every interaction with every product.

The Seattle store opened to Amazon employees more than a year ago, and has been in testing ever since. (When I stopped by in November, I was not allowed beyond the gates, and an Amazon guard wouldn’t let me take a photo.) Some of the issues that reportedly delayed opening include children moving objects around, people with similar body types confusing the cameras, and crowds. Dilip Kumar, vice president of technology for Amazon Go and Amazon Books, told the Wall Street Journal that employees didn’t just test the technology, they trained it.

Now, it seems, the market is ready to go to market. And with its debut comes dual waves of excitement and concern: excitement that the ennui of the checkout line may be obsolete, and concern that Amazon is about to eliminate the nation’s 3.5 million cashiers.

Gianna Perini, the Amazon executive who heads up the Go project, told reporters the store’s effect on employment would likely amount to a reorganization, with some jobs (cashiering) shifting to others (stocking, customer assistance). “We’ve just put associates on different kinds of tasks where we think it adds to the customer experience,” she told reporters.

That is one possibility: A company automates, makes more sales, and redistributes workers to less taxing jobs. At Panera, one of a number of U.S. restaurant and fast food chains to invest in ordering kiosks, CEO Blaine Hurst told the Atlantic that “labor has been redeployed back into the café to provide a differentiated guest experience.”

Of course that’s what managers always say. When United debuted kiosks for baggage tagging at O’Hare in 2014, the airline said it was about efficiency. But if “efficiency” doesn’t mean reducing headcount, which it often does, it means doing more business without hiring more workers. Kiosks reportedly cut check-in costs for airlines by a factor of 25!

It’s also likely that new supermarket purchase technology could eliminate low-wage, low-skill jobs and create high-wage, high-skill jobs. That’s how economists usually view technological innovation in relation to labor. This transition is good for the economy in the long run, as workers get better jobs. In the case of Amazon’s technology, a store might need fewer cashiers but more optical engineers and software developers. That expense would be covered by the increased rate of sales.

Even this upside, though, hurts the laid-off workers who don’t have the skills to be software developers. And some analysts worry this kind of improved self-check-out tech will hit retail and food and beverage jobs particularly hard, a sector where job losses from e-commerce—which disproportionately impact women and people of color—have drawn little attention, especially compared with blue-collar jobs held by white men, like coal mining and truck driving.

A final option is that Amazon’s tech may prove expensive to deploy and difficult to manage, and its arrival in Seattle is akin to the brief run of the Concorde—miraculous, revolutionary, but ultimately, too much trouble to keep up in a low-margin business like convenience stores. We don’t have any idea if the mysterious system behind Amazon Go can be distributed at competitive cost and at scale, like an iPhone, or if it is an expensive, one-off display of ingenuity, like the many gewgaws corporate America once paraded at the World’s Fair. History is littered with automats and self-cleaning houses, after all—time and labor-saving inventions that were displayed or put to market, but failed to last.