Of Course Instant Groceries Don’t Work

Obstacles include: space, time, reality.

groceries in grocery bag held by two cursor hands
Erik Carter / The Atlantic

This article was featured in One Story to Read Today, a newsletter in which our editors recommend a single must-read from The Atlantic, Monday through Friday. Sign up for it here.      

More than 20 years ago, as the rubble of the dot-com boom was still smoking, Wired magazine published an autopsy of the grocery-delivery start-up Webvan. The company had just filed for bankruptcy after evaporating the better part of a billion dollars of investment funds in about a year and a half, and the tale of its downfall opens with a sentence that, in retrospect, is pretty funny: “In the sober days of 2001, it’s hard to imagine a time when a company with an untested plan for an online grocery shopping service could inspire private investors to instantly part with hundreds of millions of dollars.”

For a while, that was true—Webvan’s failure, along with that of its contemporaries HomeGrocer and Kozmo.com, briefly made food delivery the third rail of Silicon Valley. But in the 2010s, as consumers showed interest in delivery apps such as DoorDash and Instacart, mail-order meal kits such as Blue Apron, and ultra-fast delivery from Amazon Prime, the itch to once again invent Webvan began anew. Today, entrepreneurs and venture capitalists still seem convinced that the company, which held its own grocery inventory and had plans to spend a billion dollars on new delivery warehouses before shutting down, got one thing exactly right: Going to the grocery store really is a problem that aches to be solved.

In the past few years, a spate of instant-grocery start-ups with names like Gopuff, Getir, Buyk, and Fridge No More have hoovered up billions of dollars in venture investment. They’ve spent these windfalls lavishly—opening hundreds of new warehouses, poaching executives from Amazon and Uber, adding new product categories, and hiring a standing army’s worth of new staff, all in an attempt to become the One True Deliverer of, say, a couple pints of ice cream, a six-pack, and that one ingredient you forgot to buy for the new recipe you’re trying out. The companies flourished amid a surge in pre-vaccine demand for grocery delivery, bolstered by the apparent belief that growth would continue even as the pandemic waned. But according to a story earlier this week in Bloomberg, that theory hasn’t panned out for these companies any more profitably than it did for their predecessors—some are shutting down, others are looking for cash infusions or buyers, and the ones that are trying to stick it out, such as Gopuff, are doing rounds of layoffs and closing warehouses they just paid to set up in order to preserve cash.

So far, inventing the new Webvan is proceeding in a strikingly similar manner to inventing the old Webvan. What’s fascinating about these failures isn’t their repetitive nature—lots of start-ups of all kinds fail, even if their founders and investors have hit on a genuinely good idea. What’s fascinating is that nationally scaled, near-instant grocery delivery is a bad idea, and people keep trying to invent it anyway.

For basically all of the start-ups in question, the elevator pitch is convenience. Everyone needs food and beer and toilet paper, and getting it can be kind of a hassle. Sometimes that’s because you forgot a key element of a dish you’ve already started cooking and can’t leave the house with the oven on. Sometimes it’s because you’ve gotten too stoned and are in life-threatening need of Doritos. Sometimes it’s because you just can’t or don’t want to go to the grocery store. The model “caters to millennials and college students, consumer populations that prioritize convenience over almost every other variable—sometimes even price,” according to a 2017 article on Gopuff’s early success in the trade publication Grocery Dive. As this convenience-minded cohort has aged into an even larger place in the grocery market, tech companies have invested heavily in attempts to reshape the industry in their image and gain an advantage over their fellow upstarts.

The pandemic strengthened the industry’s conviction that we’ve reached a turning point in the appetite for instant groceries. People rapidly switched their consumption to delivery services that many of them had never used before, such as DoorDash and Target’s Shipt service, and delivery apps and instagrocers began to expand their product offerings beyond convenience items and into a wider array of perishable foods to meet the sudden explosion in demand. Even as interest rates and inflation rose earlier this year, making large funding infusions difficult to secure and prompting consumers to rethink their spending, a Gopuff co-founder, Yakir Gola, told The New York Times that the eventual dominance of his company’s business model was practically fait accompli. “The world is moving toward instant,” he said in April, “and Gopuff is at the forefront of that.”

So far, the industry’s rosy outlook hasn’t actually become reality. Even established delivery start-ups such as DoorDash and Instacart, which largely skip the enormous expense of holding their own inventory in warehouses in favor of relying primarily on third-party restaurants and grocers to store and supply the food they deliver, are struggling to maintain their footing and reach profitability as people fall back into pre-pandemic habits. And DoorDash and Instacart avoid many of the functional downsides of the newer crop of apps. Pandemic or no pandemic, delivering highly perishable goods to millions of people, often with the promise that those goods will arrive in as little as 15 minutes, has proved a very tricky business: The unit economics are bad, the margins are bad, and the logistics infrastructure necessary to make the actual service function, even unprofitably, is extraordinarily complicated (bad). At base, these app-based, Silicon Valley–hyped start-ups keep running into two very inescapable IRL limitations: space and time.

Consider the inherent nature of groceries: They rot. That is, in fact, sort of their whole thing. That’s one of the primary reasons that hyper-fast delivery of, say, raw chicken thighs or salad greens or a loaf of bread might seem enticing at first glance—you don’t have to guess during your Sunday grocery trip whether you’ll be in the mood for a chicken Caesar salad on Thursday. But the risk of choosing wrong and tossing out food doesn’t entirely disappear. Instead, it’s absorbed by the instagrocer. By encouraging capriciousness, companies such as Gopuff make demand difficult to predict and create a prodigious amount of food waste, which is bad on its own terms and also very bad for a company trying to make money by selling food. In April, a former Gopuff manager told The New York Times that at least once or twice a week, he threw away thousands of dollars of perishable food immediately on arrival because he was being sent things he didn’t have any room to store.

Quick expansion makes it impossible to do the kind of market research that a more traditional grocery business might conduct before moving into a new market, which would then guide how a company stocks new stores in order to reduce waste. Those new stores also tend to be part of long-standing regional chains, which have heaps of data on when and where their shoppers buy certain things. Weekly bulk shoppers, whose habits are steadier and more predictable, mix with last-second drop-ins. Fresh foods nearing their expiration date can be marked down for quick sale or used by the store’s internal kitchens in prepared foods, which are a booming business for grocery retailers.

Then there’s the problem of real estate. To have any hope of making short-turnaround grocery ordering work, new companies have had to open up huge numbers of warehouses that are close enough to residential neighborhoods that assembling and delivering a couple bags of groceries in half an hour is broadly feasible. That means putting a large range of highly perishable products into many distinct locations. To do this in highly competitive markets like New York, some of the companies rented out neighborhood storefronts that used to sell things directly to the public, covered their windows, and used them as mini fulfillment centers or “dark stores,” but only for app orders. If you were out walking your dog after work and wanted to pick up some ingredients for pasta on the way home, you couldn’t just stop in and grab a few things. In real terms, that means these companies built huge networks of what are essentially grocery stores, occupying what is sometimes very expensive commercial real estate, but mainly for people who had forgotten something at the real grocery store, and only for people who had downloaded a special app to fix their mistake.

None of this is to say that grocery delivery as a concept is without merit, or that it doesn’t have some real and obvious upsides in some circumstances. Delivery services offer flexibility for caregivers, people with erratic schedules, and those with disabilities that make errands difficult. And the service can be successfully offered at smaller scales; FreshDirect, for example, has operated in New York City for decades. It’s also the opposite of a start-up looking to gobble up market share from traditional grocery stores. It currently has one warehouse, charges significant order fees, and requires that most orders be made a day in advance. Gopuff itself was profitable before venture capitalists took an interest in expanding the company, delivering mostly shelf-stable snacks, alcohol, and smoking paraphernalia in the areas around a handful of large universities.

According to the people leading the instagrocers that remain, all is not lost. Gopuff told Bloomberg that it still has a billion dollars in the bank, and dialing back its push into a mind-boggling number of new markets should indeed help conserve cash and buy some time to make its existing business more stable. Like Gopuff, some of these apps had genuine, profitable success at smaller scales, delivering fewer things in fewer markets before the push for exponential growth began, and there are lessons to be learned from that past stability. Inventories can be simplified and adjusted for their markets; processes can be made more efficient; prices can be raised; delivery times can be lengthened. This type of service is almost certainly possible at some scale, even if the end product cannot be as quick or as omnipresent as it is envisioned by venture capitalists.

But if even Instacart, with its low overhead costs and middleman business model, can’t be made consistently profitable under favorable business conditions, then there’s good reason to believe that large-scale rapid grocery delivery is just not possible at the scale and speed that tech investors desire. These start-ups are expanding their services to meet a customer demand that, by all indications, does not yet exist. Their biggest problem might just be that people like going grocery shopping. Not everyone, and not all the time, but it’s hardly the universally reviled task that investors and entrepreneurs seem to assume. Some regional grocery stores, such as Publix and Wegmans, have ardent fan bases. Many people genuinely prefer to pick out their own meat and produce from a selection of possibilities, or see what looks good before deciding what they’ll eat for dinner. (Delivery services, meanwhile, have to rely on stock images of, say, raw chicken or romaine hearts.) And going out into the world and having interactions with others—even the momentary kind that you’re likely to have at the grocery store—is good for people in ways that most of us instinctively understand, even if we’ve never really thought about it. When I was a kid, my mom liked going to the grocery store because she had developed a rapport with one of the cashiers, Miss Linda, over the course of years, and enjoyed going through her line.

Venture capitalists do not have much to offer these very normal people. The notion that true convenience is staying at home with everything you need brought to you, instead of living in a neighborhood where the things you need are available nearby in the course of your day, is, in my mind, a huge tell as to why investors refuse to stop losing money on these companies. It’s a consistent blind spot of the industry, and one that betrays the limited imagination with which wealthy investors envision the lives of regular Americans—if they really bother to envision our lives at all. To many billionaires, isolation away from the hoi polloi must sound luxurious and desirable—or, at least, that belief is commonly reflected in the lives they lead, the businesses they fund, and the policies they champion. But as many Americans realized during the worst days of the early pandemic, when demand for grocery delivery soared, that kind of isolation isn’t all that fantastic of a lifestyle choice. Mostly, it’s just kind of lonely.