With its flat $2 fee for 30-minute deliveries of junk food and booze, GoPuff is having a very good pandemic. Orders are up four-fold at the SoftBank-backed startup but hypergrowth comes with problems: high costs, low margins and unreliable drivers.
Rushing through their empty offices in a resurgent section of industrial Philadelphia, Rafael Ilishayev and Yakir Gola, the cofounders and co-CEOs of GoPuff, are a blur of activity amplified by the silence of the cavernous lobby. “We’re here every day,” says Ilishayev energetically. “A lot of people want to come back.”
Built from the bones of an Irish bar and a boxing gym, and spanning an entire city block, the 40,000-square-foot space was meant to be the buzzing new headquarters for the seven-year-old online delivery service. Instead it sits vacant, rows of unoccupied desks framed by blank white boards and empty pantries.
But GoPuff has hardly been idled by the pandemic. The real action is unfolding at more than 200 miniature fulfillment centers across the country —half of which were set up in the past year—where some 3,500 employees work around-the-clock shifts, jostling past one another to grab, pack and send out the overpriced impulse buys that customers in 500 cities are punching into their iPhones: a $6 pint of Ben & Jerry’s Americone Dream ice cream at 9 p.m.; a six-pack of Budweiser for eight bucks minutes before the first pitch; a new mother’s midnight call for a box of 140 Pampers Swaddlers ($34).
Founded in 2013 as a late-night service for college students to order junk food, rolling papers and condoms, GoPuff has gone mainstream, and now offers 3,000 items ranging from over-the-counter medicine and laundry detergent to pet food and nail polish—all delivered in around 30 minutes for a flat $2 fee.
“We started this business because the convenience store was anything but convenient,” says Ilishayev. Adds Gola: “The immediacy we offer is unmatched.”
Its warehouses, set up in rented spaces of 10,000 to 15,000 square feet, are the key to how it competes in the crowded, cutthroat delivery business that swelled to $66 billion even before the pandemic, according to GlobalData. Unlike other middlemen like Instacart and Uber Eats, GoPuff acts more like a retailer, buying directly from giants like Anheuser-Busch, PepsiCo and Unilever. It generates revenue not so much from the nominal delivery fee, but from slapping a hefty markup on the stuff and, increasingly, selling better placement on its app.
Business has been booming since the pandemic hit. Sales reached an estimated $250 million in 2019, and orders are up 400% this year, but it is not profitable and admits it’s still not making money on a fourth of its centers.
“We want to be the world’s go to solution for everyday needs, not just the U.S.”
Ilishayev and Gola are spending freely in an attempt to manage the hypergrowth. GoPuff has already spent an estimated $150 million of the $750 million they raised, mostly from Softbank, just 12 months ago (another $100 million was distributed to early investors). That leaves the business, which the VCs valued at approximately $2.8 billion, with around $500 million left in the bank. The company won’t confirm either its cash on hand or its burn rate and insists that they could become instantly profitable if they stopped investing in growth. Much of the money is being spent on automation, routing software and other technology to help them efficiently pack and deliver orders. “Our ambitions are large. We want to be the world’s go-to solution for everyday needs, not just the U.S. We continue to make massive, massive investments in tech to get there,” says Ilishayev.
It is not a new idea—and not necessarily a good one. Essentially the same model (with the exact same financial backer: SoftBank) was tried out in the first dot-com bubble. In 1998, two ambitious investment bankers founded a company called Kozmo.com, with the goal of delivering DVDs, snacks and drinks to what were then called “web surfers” in under an hour. They raised $250 million (about $400 million in today’s terms) and went bust in less than 3 years. Turns out super-fast delivery of low-cost groceries is a very tough way to make a buck. GoPuff says it is different because of its $2 delivery fee (Kozmo was free) and $10 minimum (Kozmo had none). It also benefits from focusing on physical clusters of customers, which pre-Covid largely meant college kids on campuses.
Still, the majority of food delivery businesses like Uber Eats (2019 losses: $1.4 billion), Grubhub (lost $18 million last year before being acquired in June), and DoorDash (reported 2019 losses: $450 million) are unprofitable despite charging outrageous markups that can reach 30% on some orders. And none of those companies have GoPuff’s additional costs associated with operating hundreds of warehouses.
Another challenge: booze. Beer, wine and liquor are among GoPuff’s best-selling products but local authorities are usually slow—and sometimes reluctant—to grant the company a liquor license; today they sell alcohol in about half their markets. Plus, delivery times sometimes stretch well past 30 minutes during peak times, in part because GoPuff depends on contract delivery drivers, who they pay up to $5 for every order delivered, and who—unsurprisingly—are not 100% reliable.
GoPuff’s mad growth also puts it in direct competition with a sea of larger rivals. “The model was beautiful when they were delivering to college students late at night. But it falls apart when you expand beyond that to be a general-purpose solution to everybody in a 50,000-person town,” says Sucharita Kodali, a retail analyst at Forrester.
GoPuff’s investors aren’t concerned. Three quarters of its warehouses now break even or turn a profit, the company claims, up from 50% at the start of the year; that includes all locations that have been open at least 18 months. “We love the model. Most importantly, the unit economics were unlike anything we’ve seen in the delivery business,” says Ryan Sweeney, a partner at Accel, noting that their $150 million investment in the last two rounds adds up to one of the three largest checks the firm has ever written. SoftBank has invested even more and it fits their signature style: throw piles of money at young, ambitious founders, and hope they use it to grow fast and dominate the market. Sometimes it works: Alibaba. Sometimes it doesn’t: WeWork.
lishayev and Gola met in a Business 101 class at Drexel University in 2011, becoming friends then roommates. Born in Russia, Ilishayev worked for his immigrant father from the age of 11, first at his sandwich shop and later at a banquet hall the family acquired. Gola grew up in a Philadelphia suburb, the son of an Israeli immigrant who ran a cash-for-gold store called “Joe the Jeweler” where he worked cleaning showcases and checking out customers; in high school, he helped modernize the business with software that tracked sales.
In college, Gola was the only one with a car and after one-too-many midnight-hour convenience store runs, he found himself wishing that he and his friends could just have snacks and smoking paraphernalia delivered to them. GoPuff was born, positioning itself as a campus app that was a “one-stop puff shop” for pot and tobacco smokers, delivering equipment like hookahs, vaporizers, grinders and rolling papers. For good measure, they also offered eye drops for red eyes and munchies like Chips Ahoy and Snyder’s Honey Mustard and Onion pretzels. Delivery stopped at 4:20 a.m.
They funded the business however they could. When they were looking for office furniture, a family friend offered what was left in an office building they were vacating; they stumbled upon four floors of cubicles, chairs and conference tables, theirs for free. They spent three months moving it into storage, taking what they needed, and listing the rest on Craigslist, netting $60,000, which they used to buy inventory and pay the Ukrainian developers who built their app.
They convinced professors to let them pitch the app to students before classes and handed out GoPuff-branded bottle openers, lighters and magnets. It was just the two of them, but they pretended to be a much larger team by using fake voices on phone calls with customers. They also exaggerated the size of their customer base to convince brands to extend inventory on credit.
“Raf and Yakir’s message to the guys responsible for opening warehouses was to get it done at all costs.”
By 2015, the business had 25,000 customers in the Philadelphia area, so during their senior year they expanded to campuses in Boston, Washington, D.C. and Austin. Within a year they had completed 500,000 orders in nine cities. They raised nearly $70 million over the next three years in three separate funding rounds.
Around this time, Gola ditched his old Plymouth Voyager and upgraded to a sleek Audi 87. Ilishayev bought a flashy BMW i8, which had doors that opened straight up like a Lamborghini. The two founders were inseparable. They shared both an apartment and an office and kept largely to themselves, according to Rob Yoegel, former director of marketing, who left in 2018. Ilishayev, the more volatile of the two, was prone to yelling at his staff. “It was his way of inspiring people,” says one ex-employee.
Everywhere they went, they looked for cheap real estate—oddball spaces like run-down strip clubs and basements with five-foot ceilings—in close proximity to customers. “Raf and Yakir’s message to the guys responsible for opening warehouses was to get it done at all costs,” says one former employee. “It was absolute chaos.”
At first, they leaned into frat-boy marketing that included ads like “Who’s getting LAYS tonight?” and “Delivery when the bra is off, but the munchies are on.” They dubbed Syracuse University the “booty call capital” because of the number of condoms ordered. Not surprisingly, they were an instant hit in many college towns.
A few cities were money pits, particularly pricey real estate markets with good late-night options like New York City, where they opened in 2015 and lasted less than two years. And there was the ongoing issue of recruiting and retaining drivers. “The number one issue in running that place was drivers,” says Kevin Moriarty, a former regional operations manager for GoPuff who came from Amazon. “It was absolutely maddening.”
Unlike Uber, GoPuff asks its drivers, none of whom are on staff or get benefits, to sign up for shifts up to eight hours, something few want to do. So when it goes bad, it goes bad spectacularly. Moriarty described a situation at Penn State. The GoPuff warehouse was crushed with 200 orders on a sunny Friday afternoon before a football game. Problem? Several drivers didn’t show up for work, stretching delivery times to as much as two hours.
The neighbors certainly don’t love them. In Philadelphia, people who live near their warehouse created a Twitter account in January dedicated to complaining about the vehicles that clogged their streets and sidewalks, the litter on the ground and the stench of cigarette smoke. In Chicago, residents voted last year against approving the permits GoPuff needed to open a warehouse.
imple growing pains, the founders say. In response to driver feedback, they have begun experimenting with shorter shifts. They don’t guarantee delivery will be under a half hour, but say their average delivery time is 25 minutes. Last year, they introduced a scanning system and an app that warehouse workers use to quickly pinpoint products. “It’s forever work. Amazon is going on 25 years and is still improving,” says Gola.
They’ve ditched their earlier raunchy marketing in an effort to attract older, more affluent customers; so far they say they have boosted their customers’ average age from 18-22 initially to 25-34. “Our number one customer in Phoenix is a retired 70-year-old Arizona State University professor. She’s ordering four times a week,” gushes Ilishayev. “These are the kind of things that really excite us.”
They are starting to do more with the valuable customer data they have, which is prized by brands of all sizes, including Hershey, which tapped them to help market its first new candy bar in over two decades by sticking free samples in GoPuff deliveries. Nightfood, a Tarrytown, New York-based ice cream startup, paid GoPuff almost $600,000 for priority placement on the app and data on who was buying its ice cream, and when.
The jury is still out as to whether GoPuff will make it or go up in a puff of smoke. The founders are in it for the long haul. “Millions of customers are requesting GoPuff everywhere,” says Ilishayev.
With the hubris of two guys who have raised nearly $1 billion before their 30th birthday, they dismiss any concerns about their ability to pull it off, not just in the U.S. but eventually around the world. Says Gola, with a wave of a hand: “We’ve done this time and time again.”