The Fantasy of Instant Delivery Is Imploding

It’s also the kind of instant gratification business that for a brief moment in recent years seemed to have the potential to change e-commerce forever. “We’d build a plan and they’d [investors would] say you’re not spending enough,” Ilishayev says. “It’s easy to get caught up when you’re having a lot of things go right.”

Gopuff is part of a class of startups that soared during the pandemic, trying to solve a logistics and math puzzle that’s dogged Silicon Valley for decades: Can an e-commerce company whisk products to your house in under an hour? And more important: Can it actually make money doing so?

Flying cars and nuclear fusion it certainly isn’t. But the problem has confounded nearly everyone who’s tried to solve it, starting in the 1990s with Kozmo.com, whose bike messengers swarmed New York and a handful of other cities, offering free one-hour delivery of everything from magazines to 16-ounce Cokes, before it was vaporized in the dot-com crash. Two decades later, Covid-19 lockdowns created the perfect conditions for the model to finally work—billions of people trapped at home, desperate to have anything and everything delivered as quickly as possible, for almost any price. Nearly $10 billion of venture capital gushed into so-called quick commerce companies like Gopuff and the Istanbul-based Getir in 2021, according to PitchBook Data Inc. That didn’t include the exponential growth of delivery apps like DoorDash, Uber Eats, and Instacart, which ferried food from restaurants and supermarkets. Meanwhile, the biggest deliverer of all, Amazon.com Inc., was notching a 40% annual growth rate.

Gopuff, which had been kicking around since 2013, emerged with a different strategy than its DoorDash ilk, mostly middlemen apps that relied on a contract workforce and often skimmed a margin of up to 40% from the restaurants and retailers they made deliveries for. Instead, Gopuff had an Amazon-like approach of storing and stocking products in its own mini warehouses staffed by full-time employees, then using contractors to deliver products to people’s doorstep for $1.95 an order. Its founders, Ilishayev and Yakir Gola, each now 29, got to know each other when they were undergrads at Drexel University, where they started an online business selling hookahs and other smoking paraphernalia to other college kids (early slogan: “Puffin’ has never been this easy”). It wasn’t until 2015 that they entered the booze business, charging an additional $2 fee per order for alcohol, and gradually expanded their assortment and ambition.

By early 2020, Gopuff had 165 warehouses covering some 600 US cities. Then, in the span of two years, the startup raised an astounding $3 billion in venture capital from the likes of SoftBank Group Corp.’s infamous Vision Fund, acquired the 28-year-old liquor retailer BevMo!, and expanded into Europe by buying two smaller competitors. By 2021 its valuation had risen to a hyperbolic $15 billion, and the founders had cashed out by selling some of their shares to investors. They bought a private plane and decamped from Philadelphia to intracoastal mansions in Miami.

But the transition from peak pandemic to a new normal has interrupted the party. Once-wary shoppers have returned to stores looking for discounts, inflation is back, and the economy has foundered. The market s of DoorDash Inc. and Uber Technologies Inc., both public companies, have fallen 67% and 33%, respectively, in 2022. Instacart Inc. pivoted toward developing software to help supermarkets run their websites and shelved plans to go public this year. Gopuff clones like the New York-based Fridge No More and Buyk went spiraling out of business; another, Jokr, withdrew from the US to focus on South America. Earlier this month, Berlin-based Gorillas, which has been desperately hunting for a cash infusion, entered into advanced talks to be acquired by Getir. Even mighty Amazon shed 40% of its mid-pandemic market cap, closed warehouses, and laid off employees.

By their own admission, the Gopuff founders never imagined this scenario. After the company burned roughly $700 million in expansion mode in 2021, in recent months it’s laid off almost 2,000 employees, withdrawn from parts of Europe, shelved grandiose plans for new categories, raised fees on customers, and halted a planned initial public offering as its valuation has plummeted. Almost 25 years after Kozmo.com’s infamous flameout, Ilishayev and Gola are scrambling to figure out if it’s still possible to crack this particular Silicon Valley obsession. Or if the billions poured into it are destined to simply .

Back before the turn of the millennium, Joseph Park thought he had the math figured out. He and a college roommate started Kozmo.com in 1998, expanded it to 11 cities, and raised around $300 million—a serious haul in those days—from the likes of Amazon, Starbucks, and, of course, SoftBank, back when it was a holding company known primarily for bringing Yahoo! to Japan.

Park’s calculus was no different from that of your average 7-Eleven manager. Kozmo, like Gopuff today, bought most of its inventory at wholesale and sold it with a retail markup, capturing the 30% to 40% gross margin of the typical convenience store. Leasing pricey warehouse space in urban neighborhoods and delivering products for free made the economics more challenging, but Park thought that building dense pockets of customers and allowing workers to make multiple deliveries on every route would keep expenses down.

There were a few problems with this. Customers tended to exploit free shipping to make relatively small purchases—say, cookie dough ice cream and a pack of smokes—which wasn’t enough for the company to recoup delivery costs. But the fatal flaw, then and now, was the belief that an online delivery service could command the same veneration and stock multiples of purer technology companies. Kozmo rode this wave of dot-com exuberance and raised capital to finance a breakneck expansion en route to going public—until it all fell apart in 2000. “We just assumed we would get the IPO done, so we had already signed up all these long-term fixed leases on warehouse space,” Park says from South Korea, where he’s now a vice president at Samsung Electronics Co. “When we couldn’t get out of them, we were stuck holding the bag. That’s what killed us.”

After Kozmo and the coinciding implosion of Webvan (a grocery service site with less emphasis on rapid delivery that failed for many of the same reasons Kozmo did), Silicon Valley treated food delivery like it was a disease. The founding of Instacart in 2012 and DoorDash in 2013 changed people’s minds by introducing the possibility that smartphones, along with the Uber-fueled popularity of gig work, made the business model more feasible.

The Gopuff founders knew little of this history. As a small child, Ilishayev (loquacious, stocky) moved to Brooklyn, N.Y., from the Azerbaijan region of the former Soviet Union with his mother; he spent his teenage years working in the local pharmacies she owned and operated herself. Gola (taciturn, slim) was born and raised in Cherry Hill, N.J., where his father was “Joe the Jeweler,” a local raconteur who operated a regional chain of cash-for-gold stores that he loudly promoted on TV ( “You got the gold, I’ve got the cash!”). Gola helped his father sell his wares on EBay, but by the time he was at Drexel, the family business was crumbling and their home was headed for a sheriff’s auction.

When they met as college freshmen in 2011, Ilishayev and Gola say, they vibed over their similar upbringings and a shared Business 101 class. There was mutual desperation, too—both were racking up student loans, and Gola was running out of money. Driving around town one afternoon for party supplies in Gola’s beat-up ’99 Plymouth Voyager minivan, the friends observed that students without cars had no reliable way to satisfy their late-night cravings. The notion of a delivery business was born.

The original Gopuff website, and the company name, referred winkingly to vaping and smoking hookahs. “We don’t judge, we deliver,” the site promised. The founders weren’t avid smokers themselves, they claim; rather, they were catering to their target demographic. “What do college students like more than anything? They like to drink and smoke,” Ilishayev says. “We didn’t have an alcohol license, so we leaned into the other vice.”

Gopuff’s early days were pretty spartan. The duo paid a Ukrainian software shop to build their website and mobile apps for customers and drivers. They rented a small warehouse via Craigslist in West Philly, then pulled out of the deal when a stranger was shot on the sidewalk outside. Products were purchased at wholesale, marked up for retail, and paid for on the co-founders’ credit cards, while they made deliveries themselves. That is, until Gola’s oil-guzzling minivan died after seven months, and he replaced it with a 2004 Dodge Grand Caravan. “Everything we had, we put into Gopuff,” Gola says.

They expanded the old-fashioned way, funneling their slender profit into expanding their operation to cities like Boston in 2014, then Washington and Austin the following year. They were unfamiliar with venture capital and were surprised when a partner from LA-based Anthos Capital LP called one day on the customer service line. He’d been tipped off by an intern who’d heard of Gopuff in college. The partner, Bryan Kelly, agreed that Gola and Ilishayev’s model could deliver items more efficiently from their own centrally located warehouses, as opposed to sending drivers to restaurants and supermarkets à la DoorDash. “I felt like these two guys would run through walls to bring the service to the world,” he says.

Ilishayev haggled out the particulars of that first $1.25 million in funding in the hallway outside his Drexel photography class, screaming into his phone like he was bargaining over the price of a faux-leather wallet. While he graduated a few months later, Gola had to drop out when he could no longer pay tuition. Joe the Jeweler was apoplectic: “Do you guys want to be delivery drivers for the rest of your lives?!”

But something real was brewing. In late 2015 the founders Googled a local lawyer familiar with arcane state beverage regulations and spent $160,000 of their new funding on a Pennsylvania liquor license. Finally they could cater to the other vice, and immediately their college clientele bought more stuff, more often. Most important, the company achieved profitability in its native city. “It’s really rare that you have a category that pushes all your key performance indicators in the right direction,” says Ilishayev. The secret sauce to instant delivery, it seemed, was the sauce itself.

Over the next few years, companies like Instacart, DoorDash, and the London-based Deliveroo raised billions in capital. Amazon, which had been considering such instant delivery models since investing in Kozmo in the late ’90s, decided it might finally be feasible. In 2014 it did a trial of a new service, called Prime Now, promising a range of household goods and food at your doorstep in one to two hours. Once again, Silicon Valley was convinced it could fulfill every impulse with the tap of a button. “As the saying goes, history is rhyming,” says Kozmo’s Park, who watched it all play out from afar.

Gopuff also got swept up in the wave, raising more than $3 billion from Fidelity Investments, Accel, and, as ever, SoftBank, which in the classic fashion of its founder, Masayoshi Son, was also backing DoorDash and playing both sides of the quick commerce market. Investors had a simple command for Gopuff: Grow. 

Although Ilishayev and Gola had proved themselves skillful fundraisers, there was little evidence they could operate a global commerce platform growing at lightning speed. They expanded to roughly 550 warehouses in 2021, opening half of them in that year alone, and hired dozens of seasoned Amazon executives, awarding them with lucrative sign-on bonuses and the promise that Gopuff had finally solved the rapid delivery puzzle. To get beyond the college crowd, which was a fraction of the mass business the company was chasing, it added an assortment of groceries and family essentials like baby food, diapers, and cosmetics.

The Amazon transplants warned their new employer that it was tricky to store and sell perishables, but the founders insisted. They wanted to give their customers a reason to delete rival apps from their smartphone. Bananas, milk, avocados, eggs, and cloves of garlic flooded into Gopuff’s warehouses, along with the condoms, day-after pills, and other old standbys that employees referred to as NOOS, or “never out of stock.” The company doubled the limits on some AmEx corporate cards to $20,000, and teams were told to replenish inventory by going to Sam’s Club or BJ’s Wholesale Club, or, when desperate, by ordering from Instacart. (The company says this kind of inventory backup occurs less than 1% of the time.)

The founders were also consumed with dealmaking. In addition to buying BevMo! in November 2020, Gopuff acquired the 23-store Liquor Barn chain in Kentucky, partnered with Uber to add Gopuff to the Uber Eats app, and acquired a startup called RideOS, for $115 million, to help manage the fleet of contract drivers more efficiently. Ilishayev, whose mother had opened those drugstores in Brooklyn, dreamed of adding an online pharmacy and began building one.

While trying to figure out its land grab in the US, Gopuff also acquired two European rapid delivery startups, Dija and Fancy. Its new European employees recall the founders giddily promising over Zoom that Gopuff would one day be worth $100 billion. “What Gopuff saw from Philly was just landmass to have our flag there. It did not seem economical,” says a former Dija executive, who says he and others warned them that European shoppers loved their local merchants. Integrating the operations of three separate startups was a “ridiculous nightmare,” adds the executive. Meanwhile, other VC-backed rapid delivery startups like Getir saturated the continent with steep discounts. Everyone was hemorrhaging money.

But Gola and Ilishayev continued barreling into the hypergrowth trap. Along with entering too many markets, they overspent on marketing, with billboards in Times Square and European soccer and Formula One team sponsorships, former finance executives say. During Gopuff’s billion-dollar funding rounds, the co-CEOs had also sold portions of their stock to investors. (Rank-and-file Gopuffers were not allowed to sell shares unless approved by the company.) After they became multimillionaires, they purchased a Gulfstream jet and mansions five minutes from each other along the Intracoastal Waterway in Miami’s Golden Beach. Gola also bought Joe the Jeweler a home in Cherry Hill to replace the one he’d lost when his cash-for-gold business went bust.

At least some of the employees who’d worked side by side in Philadelphia with Ilishayev and Gola suddenly felt shut out. Contact with their bosses was reduced to following them in press accounts or on Instagram—relaxing with friends on a Tulum beach, partying in nightclubs after Paris Fashion Week, or marveling over Ilishayev’s 15-foot-tall wedding cake at the Ritz-Carlton Key Biscayne.

In May 2021, Amazon announced it was ditching Prime Now after concluding that most customers are content with getting orders the next day, rather than paying extra for two-hour delivery. The gnarly and expensive exercise of building an entirely new set of warehouses inside cities, closer to people’s homes, just wasn’t worth it.

The Gopuff founders ignored Amazon’s move as any sort of ominous sign—if anything, it meant the biggest player was out of the way. It wasn’t until late last year when one of their advisers, former EBay CEO Meg Whitman, urged them to gird for an economic downturn that they decided to tap the brakes. The co-founders conducted a small round of layoffs in January, paused the company’s acquisition spree, and negotiated a $600 million cash infusion, the majority of which came through a convertible note—an arrangement that, unlike a traditional VC round, wouldn’t require Gopuff to reprice its once highly d stock. Ilishayev also shuttered his pet project, the pharmacies. “It just got to the point where we had too many objectives,” he says.

Despite these tactical retreats, the company appeared to be s­­pinning out of control. It operated in 1,000 cities, many of them new markets where “Gopuff” sounded more like something a pledge might hear at a fraternity party. While its older warehouses next to college towns like New Brunswick, N.J., were thriving, with 1,800 orders a day, newer facilities were struggling. For example, one in the Bronx was seeing only around 150 and losing up to $15 per order, according to a former executive. In its frenzy to stake a claim in as many cities as possible, Gopuff hadn’t considered that many of the area’s largely low-income residents relied on the Supplemental Nutrition Assistance Program (formerly known as food stamps) and lacked the disposable income to spend on marked-up cereal.

Meanwhile, the company had doubled headcount the prior year, and most executives had never even seen the inside of one of its warehouses. Gopuff employs roughly 18,000 couriers and pays them an hourly wage for shifts they sign up for, plus a flat fee for each bag they deliver. But outside some warehouses, groups of couriers often idled in their cars without any work to do. Inside, front-line employees describe sprawling chaos, with primitive computer systems, slow order volume, and waves of inventory from suppliers that didn’t correspond to actual demand. Waste was rampant, they say; mountains of moldering fruit and baked goods had to be chucked. One Amazon transplant—who like dozens of others lasted less than a year—says workers were told to put items in black bags so no one could see the prodigious waste.

The black bag trick didn’t work in New York City. Residents in one Upper West Side neighborhood complained to their city councilwoman that homeless people were gathering outside the warehouse every night to eat the leftover food. Ilishayev says the waste was the “natural cost of testing, learning, failing, iterating, and improving” in groceries, and that the company quickly dialed back on items customers weren’t ordering, such as ginger.

Employees had similar complaints in Europe: rotting food, shortages of critical products, and low customer loyalty, with shoppers attracted to the app only by a steep discount on their initial order. Over the summer, Gopuff decided to withdraw entirely from Spain and France to focus on the UK. Ilishayev admits they should have nailed one European market before expanding to the next, but says, “If the market kept going the way it was going, it wouldn’t have mattered.”

But it did: Inflation coincided with people once again getting off their couch to shop for meals and groceries, sick of dishing out endless delivery fees and courier tips. In New York City alone, Gopuff sales have fallen 27% from a high in May, according to market research firm YipitData. Smaller rivals went out of business as others considered merging. A Gopuff IPO, once in the early stages of planning with Morgan Stanley and Goldman Sachs Group Inc., was put on hold.

In meetings with their board this spring, Ilishayev and Gola promised to cost-cut their way to profitability. They closed 76 newer warehouses in the US, signaling that “instant delivery” wouldn’t be quite as instant anymore. Around that time, delivery fees went up a dollar to $2.95 (followed by an alcohol fee increase), the first time the company had ever raised its prices. They also fired 3% of their employees in March and another 10% in July. Workers traded stories of exiting colleagues being required to sign draconian nondisclosure agreements, and soon several high-ranking execs quit, leaving behind their unvested stock, which one calls “the best Exhibit A of my confidence level.” The layoffs continued this month when Gopuff abruptly fired as many as 250 people as part of a plan to outsource customer service teams overseas to save cash.

The founders insist that Gopuff is on solid footing. They say they have $1 billion (down from $2 billion in March) in the bank to ride out the economic storm, along with new ideas, like private-label products and hot foods prepared in a new line of warehouse kitchens, that will improve margins. In August, Gopuff appointed former WeWork executive Ted Stedem as Chief Financial Officer. They also think they can offer their service to competitors, and earlier this month clinched another deal, along with Uber Eats, to deliver products for Grubhub’s 32 million diners. “When you are delivering pizza, alcohol, and all your general merchandise, all in a basket north of 30 bucks, and you are batching deliveries”—sending out drivers with a cluster of orders to proximate customers— “that is how you make the economics work,” Gola says.

But the math is not all that different from what it was almost two-and-a-half decades ago when Kozmo.com tried to make it work. Instant delivery might still pencil out when you factor in things like delivery fees and extra revenue from sponsored brand placement, but once the consequences of overexpansion, excess hiring, and forgone dreams of world domination are factored in—the cost of hubris, in other words—the numbers just don’t add up.

The options for Gopuff to salvage some version of its business are dwindling. It could raise prices even more. It could also lean harder into advertising, a tactic being pursued by larger rivals including DoorDash, Instacart, and Uber Eats, which sell sponsored listings and homepage banners to brands and restaurants to drum up higher-margin revenue. In some cities, Gopuff finally cut back on its promotions and discounts and added special fees for orders under a certain size. Angela Lee, a professor at Columbia Business School who teaches courses on venture capital, is unconvinced this strategy will be successful. “Once they have a captive audience, they’ll ratchet up prices, which is exactly what Uber did,” she says. While it worked out for the ride-sharing company, she says quick delivery doesn’t have the same kind of ubiquitous demand.

The other option is to try to get those order sizes up by tapping into another burgeoning industry, one with roots in impulse buying: marijuana. Rivals are already doing it. As of this year, customers in Toronto can order weed through DoorDash and Uber Eats, and the idea recently came up as a partial joke at a Gopuff board meeting, the founders say. With legalization across the country, a compact package size, and high margins, weed would fit nicely into the company’s warehouses, those temples of intemperance. “We’re Gopuff,” one director said during the discussion. “It’s our birthright to sell it.”

(Updates with information about Gopuff’s financial standing in 36th paragraph. A previous version was corrected to remove reference to market share in 21st paragraph and to clarify that the apps are middlemen in last paragraph.)

More stories like this are available on bloomberg.com

©2022 Bloomberg L.P.