When I say that I have an obsessive fascination with cashiers, Jason Buechel, CEO of Whole Foods Market, laughs. He understands my niche (and admittedly nerdy preoccupation).
Cashiers have been the target of automation for decades, long before the mass application of generative AI. The role was first disrupted by James Ritty, a saloonkeeper from Dayton, Ohio, who invented the mechanical cash register in 1879. In 1949, inventor Joe Woodland was inspired by Morse Code to create the modern barcode, which was finally widely adopted throughout the 1980s and ’90s—disrupting the cashier profession once more. Around this time, self-checkout machines were introduced into grocery stores, leading many to wonder whether cashiers would soon become irrelevant.
Of course, as anyone who has tried scanning produce or alcohol at a self-checkout machine can tell you, cashiers are in fact not irrelevant—at least not yet.
Now Just Walk Out technology and smart shopping carts threaten to reduce the need for cashiers once more. According to the Bureau of Labor Statistics, the number of cashiers is projected to decline 10% from 2022 to 2032. But still, this will leave nearly 3 million Americans working as cashiers. The question that remains is what technology these workers will be using, and how it will impact their lives.
Companies like Whole Foods sit at the center of this question. Purchased by Amazon in 2017, the organic grocery chain’s technological “Amazonification” has reportedly included everything from cameras that follow shoppers’ every move to performance management tools that measure how many units online fulfillment specialists pick per hour.
“Technology in general, but AI more specifically, is going to fundamentally change every part of our business,” Buechel says.
Fast Company sat down with Buechel to discuss how artificial intelligence is transforming work in the grocery aisle and beyond. (This interview has been edited for length and clarity.)
How is Whole Foods embracing and utilizing AI?
One of the things that I think we have a particular advantage in is that we’re part of one of the best tech companies in the world. . . . When I look at our team members today, we have a lot of different administrative, mundane, repetitive tasks that are not exciting and don’t always add value. That’s everybody from folks working in our store-support roles on merchandising, supply chain, and operations, to folks in our stores. Today we’ve got a lot of different manual and disparate tasks that we want to be able to bring together. And we’re already finding ways in which AI is going to be able to do some of those things so our team members don’t have to. What we want to do is reinvest that time into our customer experience and our team member experience.
Let’s start there then. How is AI changing the customer experience?
On the customer-experience side, we want team members to have more time to interact with our customers. The idea is, how do we take something that’s not value-added today, where AI can help support it, and then find areas that we can reinvest that time into?
When we think about AI across our supply chain, whether it’s ordering products, to better understanding the different impacts to the supply chain; whether it’s a weather event or seasonal event that’s taking place, the question is: How do we make sure that we’ve got the best information to adjust the business accordingly? And that’s something that I’m excited about. We can improve how we forecast as a company, and we can improve our ordering tools. All of these things will then, in concert, support a better customer experience.
Some of it will be indirect. Like, as we free up time from our team members who aren’t doing those mundane tasks, they’re going to free up time [to interact with customers]. For instance, those who are apprentice graduates, whether they’re a butcher or certified cheese professional or a bakery decorator, they’re going to have an opportunity to have more connections with customers, which is, I think, one of our differentiations as Whole Foods Market.
There are also some things that will be directly impactful to our customers. When you look at our customers who are already leveraging technology—whether they’re ordering groceries online or they’re using our app or browsing on Amazon to be inspired about products that they might buy, there are ways by which we can help customers build baskets, plan their meals, and help save them time.
We’ve already got some things that we are building in today’s digital experiences. But I think this is where there’s unlimited potential. Like, I brought up the weather. Customers, based upon what the weather is like, change their buying behavior that day. Days that are cold and rainy, I’m telling you, we sell a lot more soup. And how you build a basket changes, as a customer. And so we’re finding opportunities by which our digital experiences can actually reflect what’s going on in the broader ecosystem for our customers.
When you look at seasonality changes—like what happens during holiday events, or when your kids are in school versus out of school—what you’re buying changes. A lot of the ways that we support basket-building today don’t reflect those things. But in the future, we’re going to be able to do that.
And the area that I’m most excited about—that will have a very direct impact on the customer experience but will happen sort of covertly—is geography.
How does geography impact how you expect Whole Foods stores to use AI?
Today, we assort [products] based upon geography. So if you’re going to be shopping nearby, we’ve got One Wall Street, Tribeca, Bowery, and Chelsea. All of those stores roughly have the exact same items. But the reality is that there are different customer needs.
So one of the things that we’re going to be able to do is better allocate space and adjust our assortments based upon those needs. [A great] example . . . is our Beverly Hills store and our Brentwood store: They’re really close together, but one is on a college campus, and the needs and products actually should be different.
This is one of those areas where we’re going to be able to use AI tool sets to understand how we can best curate product assortment, not just on a geographical level, which a lot of folks do, but ultimately down to the store.
We call it store-level customer centricity, but what we want to do is make sure that we’ve allocated the right space, with the right products, at the right price, across our entire store network.
To do that work today would require hiring hundreds and hundreds of team members. And this is where we’ll be able to help streamline work, which will allow our buyers to hopefully spend more time on foraging and discovering the best products, and figuring out how those go into our offerings.
You say you would have to hire hundreds of people to do this kind of work. Do you expect any of these applications of AI to impact the number of employees you have at store locations?
Right now, no.
When I give the example of hundreds of employees, I’m talking more about like, if we were going to assort differently for every store, we’d have to hire bigger category and merchant teams because today we’ve got folks who manage all of that.
At the store level, that’s where we’re really looking at how to redeploy that time spent. We think this can actually be a win for Whole Foods, a win for our team members, and a win for our customers. AI can take some of those tasks that are just administrative, which I will say, by and large, are mostly the tasks that our team members don’t get excited about. But by and large, our team members love interacting with our customers.
For instance, our apprenticeship programs: This is a way we can have more team members have the time to [participate]. So when we look at every single store in particular, I don’t think it’s fewer team members. I do think AI allows us, as we grow bigger and bigger, to be more efficient.
My aunt works at a cheese counter in a grocery store. Smell definitely seems like one of those senses technology can’t replace yet.
When it comes to cheese, if you can’t smell it or taste it, it’s hard to make recommendations.
What you said about growing bigger is really interesting, because when I speak to executives who are in growth industries, they’re more likely to say that AI will allow them to do more. And executives who are in more difficult climates seem more likely to view AI as a cost-saving opportunity.
One of the great things that we have going for us is our business is in a growth position right now, both in growing customers and sales in our existing stores—but we’re also growing a lot more stores.
We’ve got a new format, Whole Foods Market Daily Shop: The first one is opening in the Upper East Side, here in Manhattan. We also have a really large store pipeline, especially when you compare the rest of visible retail right now. Most [competitors] are contracting stores versus growing. We’ve got 75 stores in development. That number is going to keep going up and up. Within the next few years, we’ll be opening about 30 stores a year between our existing and new store format.
So a big part of this conversion is how do we make sure our team member base is in a position to help support our growth as well. And one of the things we want to do is have engaging work and careers where people feel fulfilled by what they’re doing and have a path to continue to move up.
The other big category I hear executives talk about is using AI to track things like employee performance. What specific tools are Amazon and Whole Foods using for employees right now, and how will AI impact them?
Right now we don’t have any active workstreams in that space. I think one of the areas that we’re sort of looking at, though, is using [AI] as a way to match up all the different skill sets that our team members have. You know, when building schedules, how do you make sure that you’ve covered for those skill sets and shifts?
One of the other areas that we’ve done a lot of work on is transferable job skills. Like, how do you get uploaded and upskilled in another part of the store? Being able to be more dynamic means figuring out who can help come in and support.
Take a store like Columbus Circle [where] you’ve got hundreds of team members working at any given point. The ability to have information at your fingertips to understand what skills [are] in place is valuable.
As you know, different volatility happens within a business. You might have had a late delivery by two hours. Or maybe at that store everything has to go through extra security because of the building it’s in. How do you adjust the workforce really quickly to still serve customers? I think there are a lot of things that will help optimize and find-tune how you can best leverage the workforce that you have on that given day, given the tasks that are at hand.
So you don’t think AI will impact things like units-per-hour tracking?
We’re not doing anything active in units per hour or productivity in that way. But you know, when you look at things like out of stocks, as an example, when you’re looking at differences between high-performing stores and low-performing stores as it relates to unit movement of products, it can help to understand what the different correlations are.
I’ll give a great example of one challenge that we want to help solve. A lot of our ordering steps for the next day are taking place in the early mornings. The problem is that there are weather impacts that happen throughout the day. We might actually be able to start looking at how we can make adjustments before things are picked up from distribution centers. So if a weather event has come along or something’s changed, how can we get some better data and information? The same thing can be used for store leadership.
Can you speak about the kinds of technology Whole Foods is leveraging that could replace the need for cashiers?
One of them that we are really bullish on right now is the Amazon Dash Cart, and the reason is [that] it has familiarity with customers. They’re used to pushing carts through stores, and what we’ve done is added new capabilities to it. This is where I think AI will play a role at some point—the digital experience that comes with the cart itself.
The customer could be able to say, “Okay, I bought these three items. Are there other items that are going to pair really well with them?” Or “I care about promotions on these types of products.”
Being able to not only help suggest things but also to guide our customers throughout the store based upon a list that they might prepare in advance, we [could provide] . . . the most efficient way to get in and out of the store. And obviously, simplifying things where you don’t have to stand in a line to then leave the store—that’s where things have been really positive.
So we’re going to continue to expand [Amazon Dash Cart]. And I think this [will help] customers leverage a greater way to shop that allows them to avoid having to stand in line at the end of their shopping trip.
It seems like we’re finally at a stage where a Dash Cart could replace cashiers once and for all. Is it fair to say that we’ve come to that point? And if so, how should we retrain those cashiers?
Well, I don’t think every customer is going to be interested in the [Amazon Dash Cart]. Sometimes it’s driven by the basket size. So customers who have a really big basket, they’re like, “I just want someone to help.” Some folks just like the interaction—especially in our stores.
One of the key reasons we’re doing cross-training is we do have some stores—I was actually in one today—that are doing significant volumes of self-checkout. We were sitting there watching, and there are team members available, and the customers were choosing [self-checkout] because in their mind, it’s the easiest way to get in, get out.
So part of what we want to do is make sure that in areas where we already have heavy volumes of self-checkout, that our cashiers have other roles that they can support in the store. In some cases, it’s picking online orders. It could be helping the grocery team working some backstock. It could be helping out in the produce department.
The old process used to be, how do you make sure everybody in the store can check out [customers] on the front end, so when you’re short, you can still check out. Now it’s the opposite, in fact, where we have to make sure that every cashier, if we’re not having as much volume, is also trained in other parts of our store.
Everybody always asks me, “What year will there no longer be cashiers?” And I don’t know whether that will happen. I think this is an instance where there is an aspect of human interaction, and I think that’s especially what our customers are still looking for. So I don’t see it at all at this point in the near future.