Have you watched that new video of Amazon Go, the store in test mode in Seattle where you just pick up stuff (prepared food and drinks, mostly) and walk out? I have, again and again. It’s pretty hypnotic, especially with the sound off.
The gist of it is that you scan in with an app on your phone as you walk into the store, grab whatever you want — and leave. “Computer vision,” “deep learning algorithms” and “sensor fusion” figure out what you’ve taken and charge you for it.
As lots of people have been pointing out on Twitter, this is not a new vision. IBM produced a similar, albeit funnier, commercial about a checkout-free grocery store a decade ago. Still, that store didn’t actually exist. Amazon Go does, has plans to open to the public early next year and could, according to the Wall Street Journal, be the first of 2,000 grocery stores of various formats that Amazon opens. Even if Amazon mucks this up, the ubiquity of smartphones and the increasing smartness of product labels seem to point toward somebody figuring this out sometime soon.
From a consumer perspective, it certainly looks appealing. Checkout is the least pleasant, most aggravating part of shopping — and today’s automated checkout machines generally don’t improve the experience at all. If I could buy stuff at my local grocery store without going through checkout, I might just buy more of my groceries there and fewer from its online competitor.
But what about the jobs? Almost 3.5 million people were working as cashiers in the U.S. last year, according to the Bureau of Labor Statistics. That’s about 2.3 percent of total employment. There are still jobs at Amazon Go; the video shows a woman restocking shelves and a man making a sandwich. But there aren’t any cashiers!
The mean hourly wage for cashiers in 2015 was $10.99, compared with the $23.23 average for all occupations. It’s not a great job. But it is still a job, one that doesn’t require a lot of experience or qualifications. And the ability to check out without checking out would appear to mean far fewer of those jobs will be created in the future.
Appearances can be deceiving, though. Automated teller machines reduced the number of tellers needed at bank branches, but the ATMs also made it cheaper to operate branches, so banks kept opening more of them and hiring more tellers. That’s no longer the case; banks have been closing branches, and the BLS expects the number of tellers to shrink by 8 percent from 2014 through 2024. But decades after the handing out of cash was automated there are still lots of people employed handing out cash.
Part of the reason is that they’re not just handing out cash. Tellers and other frontline bank employees have increasingly been enlisted as salespeople pushing a variety of financial products on customers. Sometimes that’s gotten out of hand, as with Wells Fargo & Co. and its bogus-account scandal. But that shift from transactions (handing out cash) to interactions (selling) is part of a broader economic trend. The McKinsey Global Institute has been tracking it for years, finding that jobs that involve interacting with other people have been on the rise even as those that are mainly about transacting or producing things decline.
In retailing, the most important trend of the 1990s and 2000s was the rise of warehouse stores and supercenters that took tons of business away from less-efficient competitors that employed more people per dollar of sales. This might explain why retail’s share of overall employment started falling in the late 1980s.
The decline seems to have stopped in 2009, though. It might have something to do with retailers rediscovering the value of in-person interaction as a driver of sales. At Wal-Mart Stores Inc. and elsewhere, there have been tentative signs of a shift from viewing retail employees mainly as costs to be minimized to seeing them as people who can help sell more stuff.
And so while the BLS projects that the number of cashiers will grow by just 2 percent from 2014 through 2024 (slower than overall employment) it figures that the number of retail sales workers will grow by 7 percent (about the same as overall employment). BLS statisticians do not possess a magical ability to see into the future, of course, but one can see where they’re coming from.
Now, it’s hard to imagine Amazon Go stores loading up on salespeople. It would be really off-putting to have somebody standing around asking, “Hey, do you want a drink with that salad?” But the rearrangement of grocery retailing that Amazon’s experiment portends is not going to be as simple as just cashiers losing their jobs.
Justin Fox is a Bloomberg View columnist.