In the early to mid-1900s, Automats – cafes where all the food came from vending machines without interaction with humans – were all the rage. They stayed that way until the advent of fast food made them obsolete. One could make the case that since there were no employees on site and all the food was delivered by machines, Automats were automated – or robotic – restaurants.

So how about a robot grocery store? No, not a store where robots shop – a store that has no human employees and also happens to be mobile: it can be located anywhere, on a city street, a rural byway or in a parking lot outside your office.

Praful Mathur and his colleagues at Bay Area startup Shotput believe they’ve created just that. The first location will start testing in Oakland in mid-August followed shortly by a second. And in September, Shotput will get its first road test at a weekend music festival in Atlanta.

Mathur lives in west Oakland. Though he’s in a nice neighborhood, “I walk outside and go, ‘Where do I go eat, where do I get groceries here?’ Instacart is way more expensive than it could be, the local grocery store is a cheap 99 cent store – people are living here! 25,000 people who need to have to access to decent food. Sprouts and Whole Foods are downtown but I would need a car or public transportation or Uber. Why is it so inconvenient?”

Food deserts aren’t just in bad neighborhoods – or cities. There are countless rural American towns where people have to drive awhile to get to a grocery store.

“Something like 25 million people in the U.S. live in a food desert. The only way you can solve this problem and do it long-term is take a small-scale SKU model – Aldi was already doing it, and Lidl, they’re known for high efficiencies, their gross margins are half everybody in the market, their contributions are double everybody in the market — remove rent, abort high overhead, take away the distributor [and] key middle men and deliver food to people faster and cheaper.”

The Shotput concept’s not that foreign — were Automats robot restaurants? (WPA Photo by Berenice Abbott, 1936)

Mathur and friends were discussing the problem one evening when somebody said the perfect solution would be to build an automated container that served as a pop-up grocery store. “We were a couple of beers in and it was kind of a joke.”

It became less funny when investors offered to back a concept if one could be delivered.

And the way to do that was with robotics.

Mathur was running a logistics and supply chain company with high-end clients like Dow Jones, Red Bull and others. He leveraged those connections to catapult Shotput into reality.

Shotput stores are containers – basically shipping containers with refrigeration components – that are completely automated inside, with rows of items in limited SKUs and a system of robotics that processes orders placed by app-wielding customers. Right now, shoppers bag their own orders as they’re fetched by robots; the next step will be to actually have the full package ready for customer pickup (shoppers still want to inspect items before bagging, Mathur says).

Financial transactions are conducted by RFID tags. Eventually drones will deliver some of Shotput’s offerings in local neighborhoods as will partnerships with services like UberEATS.

While there are certainly still bugs aplenty to work out – all things tech these days launch in beta and make real-world tweaks as they go – two test locations have been approved in Oakland, including one near Mathur’s neighborhood. Eventually there will be a need for centralized DCs and some permanent to semi-permanent locations, but for now it’s all mobile.

Limited SKUs are an obvious issue, but Mathur says that can be turned into an advantage by working with a wide variety of suppliers, rotating vendors and particular offerings.

Shotput is working with 100 brands right now and actively seeking involvement from more.

“We can change SKUs very frequently because we are a high volume store – Walmart would have a harder time doing that because of the 30,000-40,000 SKUs they have to deal with in their grocery section. Selection is the tradeoff.”

To counter that, Shotput’s app goes way deeper than any other retail app. With limited selection in each Shotput store, the app helps shoppers put together grocery lists tailored to individual dietary needs or restrictions, from losing weight to avoiding gluten. It can also keep customers within predefined budgets.

“We don’t want the lack of selection to be an issue, which is why we’re going to be creative and give people reasons to keep coming back. Otherwise, to compete with selection is nearly impossible.”

Local brands, farms (where ever local is) can partner up

What Shotput can do is partner up with local brands – where ever local happens to be – and provide them access to a national audience.

“We’re encouraging local brands and manufacturers to join our development process to grow together. One of my deep desires is not just to have local brands succeed here but to have brands from, say, Atlanta have success in California and California have success in Atlanta,” Mathur said. “One of the things we have in this country is we’re all pretty united in our ability to accept unique products and we all gravitate towards food. Politically we’re getting shifted away from each other but from a food standpoint we’re being drawn together.

Farmers will be welcome to partner directly with Shotput.  “We’d give you our standardized packaging and you’d send it to us and we’d figure out how to distribute it around the U.S. depending on need. Longer-term we’ll figure out how to do produce well. But this is how you help local producers really compete effectively on a national level or even on a regional basis.”

And while some may see Mathur’s mission as yet another assault on human employment, he says the jobs created by local brands doing more business will offset any loss of employment in areas that are currently food deserts anyway.

A Better Price Target Makes a Better Price Point

Mathur expects to be able to be highly competitive on a pricing standpoint.

“I’m able to take local brands and expand them, which means I get a better price target, which means my customers get a better price point,” he says.

Pickup locations will be in high-density areas like train stations and community centers. The first is planned in a district with high-end restaurants and bars, thus a steady stream of traffic. The idea, ultimately, “is not so much to be convenient but to intersect with you in the middle of your day. We’ll go to office parks. Once we want to do home delivery, we want to [launch] the way Walmart launched – they went to all the local neighborhoods and now 90% of the population is 20 miles from a Walmart.”

The Shotput model starts with a central distribution point in each region served with additional depots where needed to service the stores.

An initial round of fundraising netted $2 million two years ago; another round is underway.

And the model will continue to evolve – several different designs are being developed now.

“We’re excited to see where we go with this. We’re not going to keep it part of a truck bed – we’re going to have semi-permanent [structures] – in Berkeley [CA] we’re looking at permanent,” Mathur says. “In a very short period of time we’ll be able to put these grocery stores all around.”

Leave a Reply.