Self-Checkout

A few weeks ago, at a grocery store in New Jersey, Christopher Andrews reached a breaking point.

ANDREWS: I was shopping at my local supermarket, and I was pushing the cart, and I had probably between 30 to 40 items. And when I got to the checkout lane, I saw several staff and associates and they were standing around and I asked, you know, “Is this lane open?” And they said, “No, but self-checkout is open.” And I looked at my watch. It was 9 p.m. I was tired. I was ready to go home, and I thought, I’m not going to spend the next 20 minutes, 30 minutes doing this after working all day. I just walked out of the store.

If you’ve been in a supermarket in the past decade, you’ve encountered a self-checkout machine. The technology comes with a promise: instead of waiting in a long line operated by a human cashier, you can control your own destiny by scanning, bagging, and paying for things yourself. In its ideal form, it’s faster and more convenient than sending your purchases down a conveyor belt to a supermarket employee. But is it really a good deal for shoppers?

ANDREWS: It’s essentially transferring what were paid tasks that cashiers do onto unpaid customers. I think a lot of shoppers are left asking, you know, what’s in this for me?

For the Freakonomics Radio Network, this is The Economics of Everyday Things. I’m Zachary Crockett. Today: self-checkout.

*      *      *

There was a time when customers weren’t allowed to do any work at a grocery store. If you went shopping at the turn of the 20th century, you’d hand your grocery list over to a clerk. They’d collect everything for you, put it in a bag, and tally up your bill. That system changed with a store called Piggly Wiggly.

ANDREWS: So Piggly Wiggly, which was founded by Clarence Saunders in 1916, was really the first self-service grocery store.

Again, that’s Christopher Andrews. In addition to being a disgruntled shopper, he’s a sociology professor at Drew University who studies how technology affects the workforce. He’s also the author of a book titled The Overworked Consumer, which explores the history of self-service.

ANDREWS: Piggly Wiggly’s innovation was to allow customers to go through the store gathering items from the aisle, which ultimately cut the labor costs and in turn, let the store offer customers lower prices. So, at the time this was really a revolutionary change.

Self-service became the norm in grocery retail. The final step of the process, though — the check-out — still required human assistance from a cashier. Saunders wanted to change that.

ANDREWS: He had this dream of a giant robot store. And so he experimented with a store that he called Keedoozle, or key does all. And it was essentially a store comprised wholly of vending machines with the idea being that we would come in and there would be practically zero staff and it would dramatically cut down on the overhead. 

The venture ended up being ahead of its time, and it failed. But it got retailers thinking about a more automated future. That future arrived in 1986, when a company called CheckRobot introduced the first automatic checkout machine at a Kroger outside of Atlanta. For the first time, shoppers could scan their own groceries. A machine would read the barcode and announce the price in a digitized voice. The invention was hailed by one newspaper as “a revolution in the supermarket” — one that would make shopping more efficient for consumers and reduce labor costs for stores. But it took a while for the idea to catch on.

ANDREWS: It was really the 2000s when it began to take off in significant numbers. This was in the months leading up to the Great Recession. And so businesses saw self-checkout lanes as possibly a way to cut costs so they could compete with the Walmarts and the Targets.

Today, self-checkouts are used by 96 percent of major grocery chains, and they account for nearly 4 out of every 10 lanes.  For retailers, the machines don’t come cheap. A 4-lane setup runs around $125,000 — and a large grocery store usually has at least 2 or 3 of them. Installation, maintenance, and software add tens of thousands to dollars to that cost. There’s a whole economy of people who help implement these systems.

ALLOY: My name is Sara Alloy. I’m the experience retail lead for North America at Publicis Sapient.

Publicis Sapient is a digital consultancy. They help businesses, including major grocery chains, find solutions to problems using technology.

ALLOY: People — you know they either love or hate self-checkout. We like to work with our retailers to kind of take a step back and actually think about how customers want to use something like a self-checkout in the general flow of the store. They want to help customers get out of the store, frankly, through checkout as fast as possible.

After helping a retailer install self-checkout lanes, Publicis Sapient will track a bunch of metrics — to see if they’re serving their intended purpose.

ALLOY: The type of data we would track would just be: time in store, time through checkout. We would look at, you know, when you start the checkout to when you would finish, how quickly could you scan an item? What kind of errors might you encounter along the way? Do you need help from a human?

CROCKETT: And what does the data say? Is self-checkout actually more efficient than the regular checkout?

ALLOY: It doesn’t necessarily create an affordance every time.   it’s kind of up to — is the customer capable of scanning something faster than a traditional employee would be able to do?

Andrews says self-checkout lanes often create the illusion of efficiency.

ANDREWS: Self-checkout lanes are like a funhouse. They create a time warp effect because we’re so busy. When we’re doing things, time seems to move more quickly.

But in most cases, we are not, in fact, faster than trained employees.

ANDREWS: Cashiers have a lot of informal knowledge about where the barcodes are placed on tens of thousands of different product packaging. They have memorized all of the codes for the various different forms of produce and fruit and vegetables. So, while we might think of it as a very unskilled job, there’s actually a lot more to it. And, in part, that’s why cashiers are often faster than we are in the checkout lane.

One thing that might slow us down is the dreaded performance anxiety of self-checkout.

ANDREWS: When you’re in there you can feel the clock ticking. You can see the eyes of the person behind you. When we’re bagging our groceries, I think there’s this very keen sense that, you know, every second we’re taking to finish our transaction, somebody is waiting for us. And so there is this social pressure to, you know, “Hurry up! Hurry up! Figure out where the barcode is!”

But the problem isn’t just our own incompetence. One consumer survey suggested that nearly 7 out of 10 of shoppers have experienced an error or technological issue in a self-checkout lane. At some point, everyone has faced an “unexpected item in the bagging area” message — or, even less welcome —

CLIP: “Please wait. Help is on the way.”

All of these problems are made worse by the fact that self-checkout is often our only option.

ANDREWS: For shoppers with just literally a handful of items, self-checkout lanes are great. They’re a solution to having to wait in those long lines. The problem is that, because of store’s employment practices, we often see more and more people, in a sense, forced to use self-checkout lanes because there’s one or maybe not even any human operated cashiers open.

So why are supermarkets replacing friendly, efficient cashiers with expensive and error-prone machines? You might think it’s to reduce labor costs. And that’s true — in theory.

ANDREWS: In an abstract sense, if you replace six cashiers with six self checkouts and one staff left to assist customers, you just cut your labor costs at the checkout by roughly 80 percent.

But the reality is a little different. After Albertsons installed self checkouts in its stores, its average employee count went up, from 118 workers per store to 128. The same thing happened at Kroger and Target. This is partly because many grocery stores are unionized. A displaced cashier who has strong employment protections might just be assigned to another role in the store. Self-checkouts also often call for additional jobs — chief among them, security officers. Because, as it turns out, when shoppers are in charge of the scanning, they take some liberties.

ANDREWS: Shoplifting is significantly higher through self-checkout lanes. They literally turned us into a nation of shoplifters.  

That’s coming up.

*      *      *

Phil Lempert is a journalist and retail analyst who’s been covering grocery trends for more than 30 years. He runs a website called SupermarketGuru. And for years, he’s been predicting the downfall of self-checkout lanes.

LEMPERT: There’s probably, in my opinion, you know, a good 30, 35 reasons to get rid of self-checkout.

The biggest reason, he says, is self-checkout theft — known in the industry as “shrinkage.”

LEMPERT: Retailers around the country are seeing their bottom line being affected. We’re talking about a huge problem. When I talk to the CEOs of supermarkets, it’s number one on their list. The shrink factor, as more people have gravitated towards self-checkout, is now four to five times what it was pre-pandemic. A lot of people can fool that checkout.

Various studies have found that theft is far more likely to occur at a self-checkout than at a checkout run by a cashier. 1 in 7 shoppers admit they have intentionally stolen something at a self-checkout — which works out to around $10 billion dollars’ worth of goods each year in the U.S. alone. Christopher Andrews, the sociologist, says that shoppers employ various techniques to dupe the machines.

ANDREWS: One way that’s been popularly referred to as the banana trick is by entering a more expensive produce item as something like an apple or banana. In one store I went to, they told me about an incident in which they went to the back of the store and they had lots more apples than the system said they should have had. And meanwhile they were missing several of these cases of seafood. And what had happened is that because nobody was staffing the self-checkout lane, people were going through entering these cases of king crab legs and frozen lobsters as cheap produce items.

But not all self-checkout theft is intentional. The technology also leads to lots of accidental shoplifting.

ANDREWS: There’s different categories of shoplifting ranging from organized retail theft rings to what they call “walk-offs” — people who walk off with items because they become frustrated. Or, as I’ve experienced, things you forget to scan, they call “bottom of basket” or “bottom of the cart.” If you forget to bring that up and scan it and you walk out, technically you’ve just shoplifted.

CROCKETT: Wow. Okay. Duly noted.

ANDREWS: I don’t think we ever really had that discussion about, you know, what are the legal implications of using self-checkout lanes? If you make a mistake, you could be charged.

Lempert says there’s precedent for this.

LEMPERT: There was a woman at a Walmart store who checked out I think it was about 12 or 15 products, and she forgot to scan a candy bar. And she was then pulled into the back room by security, and threatened to go to jail, for not scanning, you know, a candy bar.

Retailers have experimented with various solutions to the theft problem: security guards, surveillance cameras, scales that can sense when the weight of a product is off, gates that won’t open until you scan your receipt. But customers sometimes bristle at the security measures.

LEMPERT: Consumers are really getting a feeling that the supermarket doesn’t trust them at self-checkout. And the truth is, the supermarket doesn’t trust them at self-checkout. But, you know, it’s very off-putting when you think about going into a grocery store, which is supposed to be a community place where you have a great experience with all these great colors and aromas in the produce department and so on, and then you go to leave, and there’s this big burly security agent coming over to you for the self-checkout.

And customers aren’t just frustrated about being policed. They’re tired of doing labor for the store.

LEMPERT: I hear from a lot of consumers that they’re saying, “You know, if I have to do my own checkout, if I have to do my own bagging, give me a discount. Give me some benefit for doing it.”

Last year, a bill in Rhode Island proposed giving customers a 10 percent discount — for using the self-checkout to buy 10 or more items. In 2019, Oregon attempted to pass a measure that limited grocery stores to just two self-checkout lanes. Both of these efforts failed. Some stores have decided that self-checkout isn’t worth all of this backlash. Last year, the U.K.-based grocery chain Booths removed self-checkouts from 26 of their 28 locations, citing a return to human-centered customer service. In the U.S., Dollar General, Walmart, and ShopRite have also pulled back on aggressive self-checkout expansion in certain markets. Other retailers have never dabbled in self-checkouts, and have instead leaned into their human cashiers as sort of a boutique offering in the digital age.

LEMPERT: Here in Los Angeles, there’s a chain called Erewhon. It makes Whole Foods look like a discount store with their pricing and with their offerings. Their driving force is having a great customer experience. You know, they wouldn’t put in a self-checkout for any amount of money.

The Dutch grocery chain Jumbo has so-called “slow lanes,” where human cashiers intentionally take more time to chat with customers. There may be a business case for decisions like this. A recent study found that checkouts with human cashiers lead to more customer loyalty than self-checkouts.

LEMPERT: The fact that we’re seeing major retailers experimenting with getting rid of self-checkout says to me that — if I was in the self-checkout, manufacturing business, I’d be worried.

There have been efforts to make self-checkout radically easier for customers. Amazon spent several years experimenting with a process it called “Just Walk Out.” At its brick and mortar stores, cameras and scanners detected when a product was picked off the shelf, and customers were automatically billed when they walked out of the door. Earlier this year, the company gave up on “Just Walk Out” and shifted its focus to smart carts that allow you to ring up items as you shop. There are also changes on the horizon for how products are scanned. GS1 — the organization that maintains standards for barcodes — has proposed a new product identifier that looks more like a QR code.

LEMPERT: It’s going to be able to tell where the product was made, when it was made, any food safety recall information. I mean, it’s just going to have a plethora of data. And you know what? Those self scanners that are there today — some of which have been there 10, 15, 20 years — they can’t read this new barcode. I think you’re going to see a major push for people to use their mobile devices to scan those codes, to be able to check out, to take the money right from our bank account or our credit card, and deliver it instantaneously to that retailer. 

But Christopher Andrews says that even if self-checkout technology is changing, one thing is here to stay: The exploitation of consumers’ unpaid labor.

ANDREWS: If the micro view is self-checkout line in the supermarket, I think the macro view is looking at the economy more broadly and looking at self-checkout in libraries, in retail stores, looking at self-ordering systems in fast casual dining, looking at self-check in at hotels. The technology is becoming ubiquitous. Even hospitals are starting to look at self-service technology to try to displace some of the tasks and the costs associated with that front-end labor of receptionists and people processing medical records. The idea of passing paid tasks onto unpaid consumers is just — it’s simply too seductive for businesses to not pursue. We’re busy running around doing these things. And you have to ask yourself: Are we getting anything out of this?

*      *      *

For The Economics of Everyday Things, I’m Zachary Crockett. This episode was produced by me and Sarah Lilley and mixed by Jeremy Johnston. We had help from Daniel Moritz-Rabson.

LEMPERT: I always wound up behind the father who was trying to teach their four year old how to use it. So, it really — it really has been a problem.