Part IThe Story
An Order of One
The refrigerator held a bottle of hot sauce and nothing else. In the version he gave his employees in 2021, it was Sriracha. In the version he posted on LinkedIn the week the company went public, it was simply "hot sauce," because by then the story had been told often enough that the label no longer mattered. The year was 2012, the apartment was in San Francisco, and Apoorva Mehta was twenty-five or twenty-six. The profiles disagree on the arithmetic, and he has never seemed interested in settling it. He had no car. He liked to cook. He had spent most of the previous two years starting companies, about twenty of them, and every one had failed.
"This was 2012 and we were ordering everything online except for groceries," he said much later, to the investor Elad Gil, in the tone of a man still faintly offended by the fact. "And here it is. There's a trillion-dollar category that's still stuck offline."
He wrote code for three weeks. There were two apps from the start, one for the person who wanted food and one for the person who would go and get it, because he had spent two years at Amazon learning that a delivery is two problems pretending to be one. When the software worked well enough to take an order, he had nobody to fill it. So he placed the order himself. He went to the store, picked the items off the shelves, carried them home, and handed them to the man who had ordered them, who was also himself.
"And of course," he told Gil, "I gave myself a nice tip."
It is a good line, and he knows it. He has used some version of it for over a decade, in interviews, at Y Combinator events, in the tidy origin myth that every company of a certain size eventually commissions from its own memory. It is also, read slowly, the most compressed account of the business Instacart would become. Three economic actors share one body. A customer wants food without the trip. A worker makes the trip. A company sits between them, writes the software, sets the price, and decides how much of the price reaches the person carrying the bag. On that first evening the three agreed perfectly, and the tip was generous because the man paying it and the man receiving it were the same man.
The next eleven years produced a company that delivered more than 900 million orders and was valued, for one feverish spring, at $39 billion. They also produced a federal class action over tip pooling, a mandatory service fee, a default gratuity quietly raised from zero to five percent, a senator's letter accusing the company of "tip baiting," and another senator's letter, addressed "Dear Mr. Mehta," asking him to stop treating the people who carried the bags as independent contractors. The founder who had once tipped himself would spend a long time arguing, in public and in court, about what everyone else's trip was worth.
He left the company in September 2023, on the day it went public. Later that year he sat on a stage with Gil and told the hot-sauce story again. He still ended it on the tip.
By the Numbers
The Instacart Ledger
~20Startups Mehta launched and abandoned between 2010 and 2012
2+ monthsHow late his Y Combinator application was
$39BPeak private valuation, March 2021
$9.9BValuation at the September 2023 IPO, priced at $30 a share
900M+Orders delivered before the IPO, containing ~20 billion items
10%Mehta's stake at IPO, making him the largest individual shareholder
~600,000Shoppers on the platform, most working under 10 hours a week
A Family That Followed the Wires
His father strung power across deserts. That is a slight romanticization. The record says only that the elder Mehta worked for an electric transmission line company, and that the job took the family from Jodhpur, in Rajasthan, where Apoorva was born in 1986, to Libya shortly afterward. Transmission lines carry current from where it is generated to where it is needed. A boy raised inside that job learns early that the valuable thing is often not the thing itself but the line connecting it to the person who wants it.
The family stayed in North Africa for most of his childhood. In 2000, when he was fourteen, they moved again, this time to Hamilton, Ontario, a steel city at the western end of Lake Ontario that most of the world knows, if it knows it at all, as somewhere near Toronto. That is how Mehta describes it too. In a 2018 Y Combinator interview he called it "a small town near Toronto," and he dropped the town's name for the more legible landmark, the way immigrants learn to give directions in other people's reference points.
"Growing up in a small town near Toronto, Canada, I remember waiting at a bus stop carrying bags of groceries in the freezing cold."
The quote appears in that Y Combinator Q&A and again, in paraphrase, in nearly every account of the company's founding. Bloomberg's IPO coverage put it plainly: he hated waiting in the cold at a bus stop with bags of groceries and believed the experience should have evolved by then. Picture the scene and you see a teenager who had spent his childhood in the Libyan heat, standing in an Ontario winter with plastic handles cutting into his fingers, waiting for a bus. It is the oldest image in the Instacart story, older than the hot sauce by more than a decade. It is also the one most clearly about belonging. A new arrival in a cold country, without a car, carries the household's food home by hand.
Years later, when the company needed a legal name, Mehta chose Maplebear. The maple stood for Canada. The bear came from the grizzly on the California state flag. India and Libya, the countries of his birth and childhood, do not appear in the name of the thing he built. Neither does anything else from the first fourteen years of his life. The name records only the two places where he became someone who could found a company. It is a quiet, almost bureaucratic act of self-definition, and it is still on the SEC filings.
He went to the University of Waterloo, the engineering school in Ontario that sends a steady stream of graduates south into American technology. He studied electrical engineering, not out of any special calling but because, as he told the Los Angeles Times, he wanted to understand "everything from atoms, all the way to what you see on a computer when you go to Google.com. I wanted to learn everything in between." He graduated in 2008. He tried a little of everything: four months as a design engineer at BlackBerry, a stint at Qualcomm, and, in a detail that tends to drop out of the triumphal versions, a spell at a steel factory. Hamilton is a steel town. Apparently the boy from the bus stop had worked at least briefly in the industry that defined it.
Then Seattle, and Amazon.
Twenty Companies Nobody Wanted
At Amazon, Mehta was a supply chain engineer on what his LinkedIn profile called the Fulfillment Engine. Fortune described the job in 2013 as "creating complex strings of math to find the most efficient routes for shipping packages to customers." He worked with people from AmazonFresh, the company's early and expensive attempt at groceries, and he formed an opinion of it that he would later repeat to a reporter with an outsider's bluntness.
"They were delivering your groceries along with Xboxes," he said. The theory, he explained, was to offset the miserable margins of perishable food with higher-margin electronics on the same truck. "That may seem like an interesting idea, but how many times do you need to buy an Xbox?"
He stayed two years. He learned two things there, by his own account: he liked building software, and he needed to be challenged. After two years the second condition stopped being met. "I was like, there's got to be more to life than this," he told Gil. In 2010 he quit with no idea lined up and moved to San Francisco.
"The reason why I quit my job at Amazon was because I wanted to become an entrepreneur," he said on CNBC in 2023, on the morning after the IPO. "I didn't know what my idea was going to be."
That candor deserves attention, because what followed was close to a parody of the period's startup culture, and Mehta's willingness to describe it as such is one of his more disarming qualities. He scanned the landscape for whatever was already working and built small companies alongside it. Zynga was growing fast, so he built an ad network for social games. Groupon was everywhere, so he built a Groupon for food, daily deals for lunch and dinner. Instagram had made photographing your plate a social habit, so he tried a review system for individual menu items. He spent an entire year on a social network for lawyers.
There were about twenty in all. "Unfortunately," he said, "all of them failed."
The lawyers are where the story turns. He shipped feature after feature, and the product kept not working. Eventually he identified the failure as his own and not the market's.
— Apoorva Mehta, to the Los Angeles Times, 2017After going through all these failures, releasing feature after feature, I realized it wasn't that I couldn't find a product that worked, I just didn't care about the product. When I went home, I wouldn't think about it because I didn't care about lawyers.
That is a more unusual admission than it first appears. Founders who fail usually blame timing, capital, or a competitor. Mehta blamed his own indifference. He had been treating entrepreneurship as a technique to be practiced on any industry he happened to land in, and he concluded that the technique was empty unless it was attached to something he actually felt. "I knew nothing about these topics," he said of the twenty ventures, "but I liked putting myself in a position where I had to learn about an industry." Learning, it turned out, was not the same as caring.
In later years he would give those two years a framework, borrowed from Steven Johnson's How We Got to Now. Johnson writes about the "adjacent possible," the set of ideas that become feasible at a particular moment because of changes in technology, culture, and social norms. In a 2024 talk at South Park Commons, Mehta described his search as a hunt for what was adjacent possible in 2010 and 2011. The framework is real and useful. It also has the neatness of something applied after the fact. The ad network, the food deals, and the menu reviews were adjacent possible too. They were simply adjacent to things he did not care about.
He also took a speed doctrine from the period, which he still teaches. "Iteration speed is everything," he told the South Park Commons audience. "The faster you go through the ideas, the more shots on goal that you have, the more chances you have of success." His first instruction for any new idea is to try to kill it: write down the ten best reasons it will fail, and if you cannot answer them, move on.
The twentieth shot did not come from scanning the market. It came from an empty refrigerator and, behind that, from a bus stop in Hamilton. Two years of looking at what other people found exciting ended with him noticing what he himself lacked.
The Six-Pack
The idea was right. The calendar was wrong.
It was June 2012. Y Combinator, the Mountain View accelerator that had become the default first stop for a certain kind of founder, had already chosen its summer batch. Mehta wanted in anyway, and he was clear-eyed about why: he "knew that getting in would dramatically improve the chances that Instacart would succeed." He was also a single founder, which made him less attractive to YC. And he was, as he put it in an essay he wrote for TechCrunch that August, late "by over two months."
He worked the problem the way a supply chain engineer would, looking for a route. He messaged every YC alumnus he knew and asked for introductions to the partners. Within twenty-four hours he had several. Then the replies arrived, one by one, all saying the same thing.
"No way."
The last of these came from Garry Tan, a YC partner, and Mehta read it the way a desperate man reads anything. "You could submit a late application," Tan wrote, "but it will be nearly impossible to get you in now." Mehta seized on the word nearly. "That meant it was possible!" he wrote. He put together an application and recorded a video. Several days later the answer came back: no.
Then he noticed something every partner had missed, because none of them had had the chance to see it. Nobody at Y Combinator had used the product. They had rejected a description of it.
So he opened his app and ordered one six-pack of beer, from San Francisco's 21st Amendment Brewery, a company named for the constitutional amendment that repealed Prohibition. He addressed it to Garry Tan at Y Combinator's headquarters. A driver named John (Mehta's essay gives no surname, and none has surfaced since) picked up the order and texted Mehta when it had been delivered. Half an hour later, Mehta's phone rang.
"What is this?" Tan asked.
"This is Instacart!"
The next day he faced four partners and almost an hour of questions about how the business worked and why it would survive. At the end, in keeping with YC practice, they asked him to leave and told him that if they chose to fund him, he would get a call. He walked out believing the string of rejections would continue. Ten minutes later the phone rang again.
"Hello, this is Harj from Y-Combinator. I can't believe we're doing this. We haven't let anyone in this late. Ever. But if you're interested, we would love to have you. Call me back."
The story became famous because it flatters everyone in it. The founder is resourceful, the accelerator is flexible, and the product speaks for itself. But look at how it actually happened. Mehta did not persuade Y Combinator with an argument, a deck, or a video. He persuaded it by turning a gatekeeper into a customer for half an hour, so that the partners experienced the thing instead of evaluating it. And the man who carried the beer, whose labor was the demonstration, appears in the founding legend with one name and no further role. The company's most celebrated delivery was made by someone the company's own story barely records.
Y Combinator helped Mehta raise $2.3 million. Fortune later counted $2.5 million from YC, Khosla Ventures, and others. Paul Graham, YC's co-founder, became one of the product's most loyal users. He and his wife, Jessica Livingston, used Instacart more than almost any other company they had funded. "Instacart is one of those rare products that's surprisingly great," Graham told Fortune. "You don't realize how good such a thing could be till you try it." It is the six-pack principle restated by the man who received the next best thing to the six-pack.
YC also gave Mehta co-founders. He met Max Mullen and Brandon Leonardo as the batch was wrapping up. Both stayed through the long middle years, Mehta said in 2018, working on "scaling culture" and "some of our most complex technical problems." They did the kind of work that rarely makes headlines. At the IPO, each held about 2 percent of the company. Mehta held 10. The ratio says something about the order in which people arrive at a company, and about how long the person who arrives first goes on being treated as the one who built it.
The Ghost of Webvan
Every investor who heard the Instacart pitch in those first years heard it with a ghost in the room.
Webvan had been founded on the same premise, groceries ordered online and delivered to the door, and had become the defining failure of the dot-com era. It spent more than $800 million in three years building warehouses and fleets, and collapsed in 2001. It was, as Elad Gil put it, "one of the most famous craters in all of Silicon Valley history." For a decade afterward, "online grocery" functioned in venture capital less as a category than as a warning.
Mehta's favorite story about this comes from a pitch meeting reported by Fortune in 2015. Partway through, a venture capitalist left the room and came back holding a floppy disk that contained the Webvan business plan. He handed it across the table.
"Go home and study it," he said.
Mehta says he would have, if he had known how to open a floppy disk.
The joke is generational. The disk is a relic, the founder too young to read it. But it also carries his whole argument about Webvan in a single object. The plan was stored on a medium that no longer worked, written for a world that no longer existed. Webvan, as Mehta saw it, had not proven that grocery delivery was impossible. It had proven that grocery delivery was impossible in 1999, with 1999's tools: no smartphones, no GPS in every pocket, no expectation that you could summon a stranger with an app. "As a result of those three things being true," he said years later, citing smartphone penetration, nearly universal card payments, and the normalization of online shopping, "Instacart was actually possible in 2012."
His answer was to own almost nothing. "Instacart is entirely a software company," he told a Startup School audience in New York in 2014, "which means that we don't actually have any warehouses, no trucks, and we don't hold any inventory." The stores already existed: Safeway, Whole Foods, Costco, Trader Joe's, the independent markets. So did the cars, owned by the shoppers. Instacart would be the line between them, routing several orders through a single trip, guiding shoppers to the aisle and shelf, and planning routes around traffic, weather, sports games, and city construction. "Instacart was operational," he told Fortune, "when it was just me working from my apartment."
W
Three Theories of Grocery Delivery
Webvan and AmazonFresh built infrastructure. Instacart rented someone else's.
| Model | Physical footprint | Mehta's critique |
|---|---|---|
| Webvan (collapsed 2001) | Warehouses and fleets; $800M+ spent in three years | Right idea, wrong technological moment. |
| AmazonFresh (2013 expansion) | New warehouses near cities; groceries alongside electronics | "How many times do you need to buy an Xbox?" |
| Instacart (2012) | No warehouses, no trucks, no inventory; existing stores and contracted shoppers | Software as the only asset. |
The most pointed vindication came from Sequoia Capital. Sequoia had backed Webvan, and, as Mehta recounted in an October 2023 conversation, Michael Moritz had sat on Webvan's board while it failed. A firm with that particular scar invested in Instacart anyway, which tells you either how convincing the first-principles case was or how badly Sequoia wanted to avoid another kind of error. Mehta has a name for that error: the failure of omission. These are the companies investors pass on because the lesson of an earlier failure has hardened into a rule that no longer fits the facts. By the IPO, Sequoia owned 14 percent of Instacart, more than anyone else.
The skepticism outlasted the early rounds. In his July 2021 letter announcing that he would step down, Mehta described those years with something close to grievance, still sharp after a decade: "We were rejected by investors who told us that bringing the grocery industry online would prove to be impossible. Retailers were reluctant to work with us because they did not believe a technology startup out of San Francisco would understand the nuances of the grocery business." He turned that defensiveness into a company value, which he described to Fortune in 2017 as an inheritance from his two role models, Steve Jobs and Jeff Bezos, his former boss. The value was called "Of course, but maybe."
"It sort of emphasizes the fact that of course this is the way things have been done before," he said, "but maybe if you use new technologies or look at the problem in a different way, you can come up with a solution that's much, much better."
Fortune's 2013 headline put the ambition more bluntly: This ex-Amazon engineer wants to outsmart Jeff Bezos. He was twenty-six and running a seven-month-old company with about two hundred contracted shoppers in the Bay Area.
One of Every Single Item
Trader Joe's had no website catalog and no API. For a company whose only asset was software, that meant Trader Joe's did not exist.
So the team drove to a Trader Joe's and bought one of everything.
Then they carried it to a studio and photographed each item, every jar of cookie butter and every bag of frozen dumplings, and entered them into the system by hand. In Mehta's 2021 letter, the catalog becomes a founding sacrament: "Because no retailer would partner with us in the early days, our team bought one of every single item from grocery stores, took it to a studio, and photographed everything to build out our first catalog." Internally the practice had a name, "ninja shopping." The company that would later reach more than 98 percent of households on food assistance began with a few engineers pushing overloaded carts through a grocery store, building the store's inventory without the store's permission.
"Y Combinator encourages startups to do things unscalably," Mehta told the Startup School audience in 2014, in what became one of the clearest field reports on Paul Graham's famous doctrine. "I find that this is one of the biggest competitive advantages that a startup has over a larger company." Early customers could place orders even when no shopper was online to fill them. "Of course," he said, "this meant that I would drop everything I was doing and fulfill the order myself." Forbes reported that in the earliest days Mehta made deliveries himself by Uber. Even after he had a company, he still had no car.
The man who once tipped himself was still, in effect, his own shopper. Something real was transmitted that way. Years later Instacart required every corporate employee to work as an in-store shopper in their first week. "It's a great way to feel connected to our customers," Mehta said in 2018, "and for our engineers to learn the ins and outs of the product quickly." The founding act became a rite of passage. Every new hire would, for a few days, be John with the six-pack.
"Using the same unscalable techniques," Mehta said, "we added Whole Foods, Costco and many other stores, and we realized we had the best product in the market." With the best product came growth. Instacart expanded to ten cities and sustained 10 percent week-over-week growth for twenty consecutive weeks. In August 2013 it launched Instacart Express, a $99 annual membership that waived the delivery fee on orders over $35. In November 2013 it added alcohol. By the fall of 2014 Mehta had more than a thousand shoppers in fifteen cities and ninety employees in a new office in the South of Market neighborhood. In September of that year he signed a partnership with Whole Foods. Walter Robb, the grocer's co-CEO, explained the decision with a grocer's caution: "Customers today are interested in options and services outside the four walls of the store." In December, Instacart raised $220 million at a valuation of around $2 billion.
Around then Mehta gave Fortune a glimpse of what that pace cost him. He no longer had time to play soccer or cook. "I've been known to make a mean grilled-cheese sandwich," he said. "But now there's a 50% chance it may be burnt."
The irony is too neat to ignore. The company had started because a man who loved to cook couldn't get groceries. Two years later he had built a machine that could deliver groceries to most of the Bay Area, and he no longer had time to cook them.
He had also started to burn out, regularly enough to notice the pattern. "I realized a few years back that I used to burn out every few months, and it was happening increasingly," he told Fortune in 2015, when the magazine put him on its 40 Under 40 list beside Taylor Swift and Sam Altman. His fix was an engineer's fix, a hard constraint: "Taking one mandatory day off every week has done wonders for my productivity." That same year, Forbes ranked him fortieth on its list of America's richest entrepreneurs under forty, worth an estimated $400 million. He had scraped onto a list of the very rich in last place, which may be the most Mehta position imaginable. In January 2015, Instacart had been named America's Most Promising Company.
He talked about leadership the way people talk about a language they picked up by accident. "Not at all," he told Fortune's Susie Gharib in 2017, smiling, when she asked whether he had thought about leadership in those early days. "I never thought of myself as a leader. I was just doing what was required." By then he was thirty, the company was valued at more than $3 billion and operated in forty-one markets, and he had just raised nearly half a billion dollars from Andreessen Horowitz, Sequoia, and Kleiner Perkins. What he had learned, he said, was introspection. "As a founder and CEO, your company tends to reflect your personality." The task was to know which parts of yourself to amplify and which to hire around.
That last sentence turned out to be a prophecy, about four years early.
The Thermonuclear Bomb
On June 16, 2017, Amazon announced it would buy Whole Foods Market for $13.7 billion. Whole Foods was Instacart's biggest partner. Its new owner was Mehta's former employer, the company whose grocery strategy he had publicly mocked and whose founder he had named as a role model. In one transaction the largest store in Instacart's network became part of the company Instacart had been built to outmaneuver.
Two years later, on a stage at the Rosewood Sand Hill hotel in Menlo Park, Jessica Lessin of The Information was still asking him how he had found out. The question lasted that long because the moment had been that close to fatal.
He chose his metaphor carefully.
— Apoorva Mehta, on Amazon's acquisition of Whole FoodsIt really was like a thermonuclear bomb against the entire grocery industry. When we look back, that may have been a turning point for Instacart.
The important word is industry. The obvious reading, the one the press adopted, was that the bomb had landed on Instacart. "People said it was the beginning of the end for us," Mehta wrote in 2021. His reading was that it had landed on every supermarket chain in North America. Overnight, Kroger, Costco, Albertsons, Publix, and Wegmans each had to face a single fact: Amazon now owned a grocer, and had every reason to learn the business from the inside. Each of them needed a way to sell online that did not depend on Amazon. Most had none.
Instacart was the obvious answer, a company that already knew their shelves (sometimes, thanks to ninja shopping, better than they did) and owned no stores that competed with them. "Our teams visited and signed every major retailer in North America," Mehta wrote, "which dramatically accelerated our business." That November Instacart crossed into Canada through a partnership with Loblaw in Toronto and Vancouver. By May 2018 it operated in eleven Canadian markets. In January 2018 it bought Unata, a Toronto company that built white-label e-commerce for grocers, for $65 million, and later announced a Toronto R&D center expected to employ more than three hundred people. The boy from the bus stop near Toronto was now building a research operation in the city he used to describe as the place near where he lived.
The Whole Foods partnership in the United States formally ended in May 2019, taking several hundred shopper jobs with it. By then Instacart had grown past it. In a CNBC interview that spring, Mehta recited the numbers: 20,000 stores in 5,500 cities, more than 300 retail partners. He also gave the quote the company's own LinkedIn account later featured: "I am singularly focused on Instacart. We are heads down, thinking about the problems we want to solve." The phrase singularly focused would come back three years later, when he used it to explain why he was leaving.
Inside the company, the shape of the business had changed. In the 2018 Y Combinator interview he described a structure far more elaborate than "a personal shopper with an app." Instacart, he said, brought together "four parties: customers, retailers, shoppers and advertisers." For retailers it built software to bring inventory online and set up stores for efficient picking. For shoppers it built dispatch, batching, routing, and "localized supply/demand matching models." For the several thousand consumer brands advertising on the platform, it built ad units and measurement tools. By then Instacart could reach more than 70 percent of U.S. households and more than half of Canadian ones. It served markets, as he put it, "as big as Manhattan and as small as Anchorage."
The fourth party, the advertisers, would end up saving the company after he left. In 2019, though, he already understood what grocers had always known and startups were slow to learn: the margin in food is thin, and the real money sits in the shelf space beside it. The cereal brand will pay to be seen. The engineer who once mocked Amazon for bundling Xboxes with groceries had found his own higher-margin item to load onto the truck. This time it weighed nothing. It was attention.
Much later, in February 2024, after Mehta was gone, Instacart announced that customers across Canada could once again order from Whole Foods Market through its app. The thermonuclear bomb had become, in one country at least, a returning customer.
Who Carries the Bag
In the beginning, the person who carried the bag was Mehta. Then it was John, with the six-pack. Then two hundred people in the Bay Area, a thousand across fifteen cities, four thousand across eighteen, ten thousand, tens of thousands. By the spring of 2020, Business Insider reported, it was 500,000.
Nearly all of them were independent contractors. That classification, more than any algorithm, was what made the model asset-light: Instacart owned no trucks partly because it did not employ most of the people driving the cars. The arrangement drew challenges almost immediately. In March 2015 a class action argued that Instacart's personal shoppers should be classified as employees, alleging eighteen violations, including improper tip pooling and failure to reimburse business expenses. That June, Instacart began offering some shoppers in Chicago and Boston part-time employee status as in-store shoppers, and the next month extended the option to Atlanta, Miami, and Washington, D.C. In March 2017 it paid $4.6 million to settle the suit.
The tip kept coming back. In April 2018, after a long controversy over an optional 10 percent "service amount" that many customers had mistaken for a gratuity, Instacart replaced it with a mandatory 5 percent service fee, restored the tip option at checkout, and raised the default tip from zero to five percent. At the end of that year it changed how shoppers were paid. The Chicago Tribune ran a headline quoting a shopper: "We're being mistreated." Business Insider reported shoppers saying the frustration was slowing deliveries. In 2019 the company introduced an on-demand option so shoppers could work more flexible hours. Mehta's own phrase for what the platform offered was "flexible earnings opportunities." Nearly every account of the arrangement used some form of the word flexible, and each meant something slightly different by it.
In February 2020, fifteen in-store shoppers at a store in Skokie, Illinois voted to form a union. Fifteen people. It was a tiny number, nothing like the scale of the platform, and it was the first union election of its kind at the company.
Then came the pandemic, and the volume of orders turned every one of these tensions into front-page news.
"The last few weeks have been the busiest in Instacart's history," Mehta wrote to shoppers that March, in a letter that called them "household heroes." He listed what the company was doing: health and safety guidelines developed with "a panel of health, food safety, and disease control experts," cleaning supplies, sick pay for in-store shoppers, up to fourteen days of extended pay for anyone diagnosed with COVID-19 or placed under mandatory quarantine, expanded ratings forgiveness, mobile checkout, and a "Leave at My Door" delivery option. He announced that Instacart would bring on another 300,000 full-service shoppers. "As more people look for immediate, flexible earnings opportunities during this time," he wrote, "we hope that Instacart can be an additional source of income."
The shoppers organized a walk-off. Their organizers' line, quoted back to Mehta in a letter from Senator Elizabeth Warren dated April 1, 2020, was blunt: "workers should not be risking their lives for pocket change." Warren's letter, addressed to Mehta at 50 Beale Street, Suite 600, asked him to "reclassify your delivery workers as employees, rather than independent contractors." Shoppers reported that the sick-pay policy required a positive test or a government-ordered quarantine, a bar CNET's headline called "artificially high." Safety kits arrived slowly; in mid-April Wired reported that shoppers were "still waiting." In June, under an agreement with D.C. Attorney General Karl Racine, Instacart broadened sick leave to cover shoppers diagnosed by a doctor or with a sick household member, and added telemedicine. On May 28, Senator Brian Schatz wrote to the company about what his office called "tip baiting": customers promising large tips to get their orders picked up, then cutting them after delivery. The tip had become a subject of congressional correspondence.
In January 2021 Instacart announced a $25 stipend for shoppers who got vaccinated. The same month it said it would lay off nearly 2,000 in-store employees, because, the company explained, grocers were increasingly using their own staff to fill Instacart orders. The cuts included the shoppers in Skokie, the company's only unionized workers.
None of this is the whole story of what Instacart did for people during those months. Mehta's letters were not empty. The company did become essential infrastructure for millions of households, including people who could not or should not have entered a grocery store in the spring of 2020. Order volume grew by as much as 500 percent. Customers spent up to 35 percent more per order. The person carrying the bag was, very often, keeping someone else alive.
Still, the arc is hard to miss. The founding story ends with a man tipping himself generously. The company's growth story includes a decade of litigation and legislation over whether the tip, the fee, and the classification fairly paid the people carrying the bags, and over who got to decide. Mehta himself had once been the boy at the Hamilton bus stop. In a sense, every shopper was doing what he had done at fourteen: carrying a household's food through the cold. The difference was that each order now had three parties, and only one of them wrote the software.
The Founder Who Hired His Replacement
In June 2020, Forbes put him on its billionaires list. A funding round had taken Instacart's valuation from $7.9 billion to $13.7 billion, and his roughly 10 percent stake was worth about $1.2 billion. He was thirty-three or thirty-four, depending again on whose arithmetic you trust. By March 2021 the company was valued at $39 billion, the third most valuable startup in America according to PitchBook, and Mehta's fortune peaked, on paper, at around $3.5 billion. Revenue for 2020 was roughly $1.5 billion. For the first time the company recorded something close to its founding promise: millions of people treating groceries as something you ordered.
On July 8, 2021, he announced that he would stop running it.
The letter is a remarkable document, both for what it says and for what it leaves out. It opens with ten years of scars: investors who said the idea was impossible, retailers who distrusted a San Francisco startup, the reporters who wrote Instacart's obituary when Amazon bought Whole Foods. It mentions the Sriracha. Then it turns to the woman taking his place.
— Apoorva Mehta, letter to the Instacart community, July 8, 2021While I had never previously contemplated the idea of someone else running the company, this changed when I met Fidji.
Fidji Simo was thirty-five, from the South of France, and had spent a decade at Facebook. In Mehta's letter she had been "a part of every big moment in the company's history," helping it grow from 1,000 employees to nearly 100,000 and leading "the world's most popular app." Fortune found her in a $5 million villa in Carmel Valley, bought during the pandemic, with moss-covered roofs that looked transplanted from Provence and a vegetable garden that included a patch of carrots. Instacart's logo had become a simplified carrot. "They're very appropriate for Instacart, no?" she laughed. Her biography rhymes with his in its own key: another immigrant who came to California, made a fortune there, and built herself a small copy of the place she had left.
The official version was a founder recognizing the limits of his skills, which is exactly what he had told Fortune in 2017 a good founder should do. Other versions emerged. Fortune reported that while Simo was being named CEO, Mehta had been "quietly pitching DoorDash on buying his company," and that the effort had failed. When it became public, it raised questions about the company's prospects. Business Insider's headline was harsher: Instacart CEO Apoorva Mehta Replaced After Chaotic Product Management. Mehta never conceded the framing. "A lot of people have said that perhaps I was pushed out of the company," he told Forbes in September 2023. "The reality is, if I wanted to be the CEO of Instacart, I would be the CEO of Instacart."
Both things can be true. A founder can decide on his own to step aside and still be stepping aside from a company in trouble. As consumers returned to stores, Instacart's pandemic growth cooled. Sucharita Kodali, a Forrester e-commerce analyst, delivered the sharpest verdict in the Fortune profile of Simo: "These companies all think that they're going to be the next Amazon. But really, they're all positioned to be the next Groupon."
Groupon. Ten years earlier, one of the twenty failed companies had been a Groupon for food. The analyst could not have known it, but the insult reached back to the very first false start and set it beside the eventual success. The company that came out of his search for what he cared about was being compared to the fad he had copied when he cared about nothing.
During 2022, with inflation rising and the virus fading, Instacart cut its internal valuation three times, ending near $13 billion in October. In July Mehta announced on Twitter that he would also give up the chairmanship once the company went public.
"So why leave?" he wrote. "Since I transitioned from CEO to Executive Chairman a year ago, I realized that I want to pursue a new mission and I want to do it with the same singular focus that I had while building Instacart. Stepping off the board will allow me to do just that."
In 2019, singularly focused had been his reason to stay. In 2022 it was his reason to go.
What It Means on Day One
Instacart, incorporated as Maplebear Inc., priced its initial public offering on Monday, September 18, 2023, at $30 a share, valuing the company at about $9.9 billion and raising roughly $660 million. On Tuesday the stock opened on the Nasdaq under the ticker CART, rose more than 40 percent during the day, and closed at $33.70. On Wednesday it briefly dipped below the offering price and closed at $30.10. In two days the market had brought the company back to roughly where the bankers had priced it, about a quarter of what venture investors had paid in the spring of 2021.
Mehta sold about 700,000 shares in the offering, worth roughly $21 million. He gave up his board seat. Fortune put his remaining fortune at $1.3 billion; Bloomberg and CNBC said $1.1 billion. Either way, he remained the company's largest individual shareholder, behind only Sequoia at 14 percent and D1 Capital Partners at 13. His eleven years at the company ended in the same week the company became public property.
On Wednesday morning he went on CNBC's Squawk Box and, asked about the drop from the peak valuation, gave an answer that sounded rehearsed and was also probably sincere: "What matters is how Instacart performs over the next few years, rather than what it means on day one."
On LinkedIn he wrote something more personal. He thanked the team, "the hard-working community of shoppers," the customers, the retailers, the brands, and Simo. Then he wrote a sentence that appears in almost none of his earlier interviews, which are dense with unit economics and four-sided marketplaces:
"Finally, none of this would have been possible had my parents not made countless sacrifices as we immigrated from India to Libya and then to Canada – all so that one day my brother and I might have a chance to pursue our dreams."
Elsewhere in the same post he called Instacart "the company I founded in my kitchen." It was a modest exaggeration. The Sriracha was in the kitchen; the code was presumably written somewhere nearby. The sentence about his parents, though, was the first time in a decade of public storytelling that he placed the whole route on the record: Jodhpur, Libya, Hamilton, and then, unstated because everyone could see it, San Francisco. Maplebear, the corporate name, had always recorded only the last two stops. The founder's farewell finally named the first two.
By then he had already started over. In November 2022, The Wall Street Journal reported that Cloud Health Systems, his new company focused on chronic illness, had raised $30 million in a round led by Thrive Capital at a $200 million valuation. Its total funding reached $42 million, with Andreessen Horowitz and General Catalyst also participating. Later reports described a further venture called Abundance, an investment firm in Palo Alto that uses AI agents to research and execute trades and was reported to have raised $100 million. The man who in 2010 had quit Amazon "to become an entrepreneur" without knowing what his idea would be did not, apparently, intend to stop being one.
The company carried on without him, and the shape it took was close to the one he had sketched for Y Combinator in 2018, except that the fourth party had become the most important. In March 2022 it launched the Instacart Platform, a suite of advertising, delivery, inventory, and e-commerce tools sold to the grocers themselves. It became, as Wikipedia's dry phrase puts it, "a broader technology platform for grocery retailers." In 2024 it earned $457 million on $3.38 billion in revenue. Its catalog spanned nearly 100,000 stores from 2,200 retailers. It ran its first Super Bowl ad in 2025. That May, Simo left for OpenAI, and on August 15 a longtime business executive named Chris Rogers became CEO. In December 2025 the company agreed to pay $60 million to settle Federal Trade Commission claims that it had misled customers about its Instacart+ membership and "free delivery" offers, while denying any wrongdoing. That was the old argument about fees and what the customer actually pays, in a new form. In April 2026 it bought a fulfillment platform in Bogotá.
About 600,000 shoppers were on the platform, most working fewer than ten hours a week.
On December 1, 2023, a little more than two months after leaving the company, Mehta sat down with Elad Gil to talk about his second act. Gil started where everyone starts, with the origin. Mehta told it the way he always does: the supply chain job that did not fulfill him, the move to San Francisco, the twenty companies, the ad network for social games, the Groupon for food. The apartment. The bottle of hot sauce in an otherwise empty refrigerator. The three weeks of code. The first order, placed by him and filled by him, which he carried home and delivered to the only customer Instacart had.
"And of course," he said, "I gave myself a nice tip."
Part IIThe Playbook
What follows are twelve principles drawn from Mehta's eleven years at Instacart. They come from the decisions themselves, from the record of what he did when the obvious move and the right move came apart, and only secondarily from his commencement-style advice. Some are tactical. A few are warnings. Taken together, they describe a founder whose most durable skill was routing, of goods and of himself: finding the shortest path between a need and the person who could meet it, then building software to make that path repeatable.
The principles
- 1.Audit your failures for apathy, not just viability.
- 2.Mine the friction you've carried longest.
- 3.Interrogate the famous failure for what has changed.
- 4.Own nothing someone else has already built.
- 5.Turn the gatekeeper into a customer.
- 6.Be your own first shopper.
- 7.Read the existential threat as a recruiting event.
- 8.Find the weightless cargo.
- 9.Put everyone on the shop floor.
- 10.Price the human layer before someone prices it for you.
- 11.Hire your successor while you still can.
- 12.When you leave, leave.
Principle 1
Audit your failures for apathy, not just viability
The usual autopsy of a failed startup asks whether the market existed, whether the timing was right, and whether the money ran out. Mehta's autopsy of the social network for lawyers asked a different question, and it is the more useful one. He had shipped feature after feature and could not make the product work, and he eventually traced the cause not to lawyers but to himself: "When I went home, I wouldn't think about it because I didn't care about lawyers."
The distinction matters because apathy looks like a market problem. A founder who does not care about his users stops noticing the small signals of what they want. He iterates on the product instead of on his understanding of the people using it, and then concludes the market was too small. Mehta's twenty failures were not wasted; he treated them as tuition in prototyping, in strategy, and in managing his own psychology. But the most valuable lesson was the one he could only learn by failing at something he did not care about.
The test is simple and uncomfortable. When you go home at night, do you think about the customer?
Tactic: After any stalled project, write down whether you thought about the user's problem in your off-hours. If you didn't, treat that as the root cause before blaming the market.
Principle 2
Mine the friction you've carried longest
For two years Mehta scanned the landscape for whatever was booming, Zynga, Groupon, Instagram, and built next to it. None of those ideas lasted. The one that did came from two old, private discomforts: a teenager at a bus stop near Toronto holding grocery bags in the freezing cold, and a man in San Francisco with no car and nothing in his refrigerator but hot sauce.
Two features of that friction made it useful. It was personal, so caring came automatically. And it was universal, so it could become very large. "Every household needs to shop for groceries," he said in 2018, "generally more than once a week." The best founding problems sit where a personal grievance meets a category that every household runs into. Mehta's search succeeded when he stopped looking at other people's growth curves and looked at his own refrigerator.
There is also a lesson in how long the friction had been there. The bus stop came more than a decade before Instacart. Founders often treat their oldest frustrations as just the way things are. Sometimes the best idea is the one you stopped noticing years ago.
Tactic: List the chores you have resented for more than five years, then mark which ones every household also does weekly. Start there.
Principle 3
Interrogate the famous failure for what has changed
Webvan was the reason nobody would fund online grocery delivery, and it became the reason Mehta could. A venture capitalist handed him the Webvan plan on a floppy disk. Mehta's response, joke aside, was that the disk was the argument. Webvan had been built for a world without smartphones, without GPS in every pocket, without the habit of buying everything online. The category had not failed. Its moment had.
The broader principle, which Mehta calls avoiding "failures of omission," is that a famous failure tends to become a rule, and the rule outlives the conditions that made it true. Sequoia, whose partner had sat on Webvan's board, was willing to examine that rule. By the IPO Sequoia was Instacart's largest shareholder.
Mehta's working tool here is the "adjacent possible," taken from Steven Johnson's How We Got to Now. For any idea that once failed, ask which enabling conditions have appeared since. In Instacart's case he named three: smartphone penetration, card payments, and comfort with online shopping. If the list is long, the old failure is evidence about the past, not a forecast.
Tactic: When an investor cites a famous predecessor's collapse, answer with a specific list of the technological and behavioral shifts since that collapse, and show how each one changes the cost structure.
Principle 4
Own nothing someone else has already built
"No warehouses, no trucks, and we don't hold any inventory." Mehta said it to a Startup School audience in 2014 almost as a boast, and it was the core of the strategy. Webvan built a parallel grocery system and went bankrupt. AmazonFresh built warehouses and put Xboxes on the trucks to pay for them. Instacart used the stores that already existed, the cars that shoppers already owned, and the shelves that grocers had already stocked, and wrote the software that connected them.
This is the father's trade applied to retail. A transmission line does not generate power or consume it. It carries current from where it is to where it is needed and takes its value from that position. Instacart's early economics worked because it was "operational when it was just me working from my apartment." Fixed costs stayed close to zero, so each new city cost little more than a set of shoppers and some catalog work.
The cost of this design came later and is covered in Principle 10. Owning nothing also means not controlling a great deal, including the classification of the people who do the physical work.
M
The Mehta Decision Timeline
Key routing decisions from first order to final exit.
2010
Quits Amazon with no idea; moves to San Francisco and starts ~20 companies.
2012
Codes Instacart in three weeks; fills the first order himself; talks his way into YC with a six-pack.
2013
Launches the $99 Instacart Express membership and alcohol delivery.
2014
Signs Whole Foods; raises $220M at a ~$2B valuation.
2017
Amazon buys Whole Foods; Instacart signs every major North American grocer; enters Canada via Loblaw.
2018
Acquires Unata for $65M; introduces mandatory 5% service fee; describes a four-sided marketplace.
2020
Pandemic surge; plans to add 300,000 shoppers; becomes a billionaire.
2021
Peak $39B valuation in March; names Fidji Simo CEO in July.
2022
Announces exit from the board; founds Cloud Health Systems.
2023
IPO at $9.9B; Mehta departs on listing day.
Tactic: Before building any physical asset, list every existing asset in your category that you could route through with software, and build only what is left over.
Principle 5
Turn the gatekeeper into a customer
Every YC partner said no to a description of Instacart. None of them had used it. Mehta's insight in June 2012 was that his application had been rejected on the merits of a pitch, and the pitch was not the product. So he stopped pitching and delivered a six-pack to Garry Tan. Thirty minutes later the gatekeeper was calling to ask what had just happened.
This applies far beyond accelerators. Investors, enterprise buyers, and partners evaluate abstractions by default, because that is cheaper than experiencing the thing. A founder whose product is better experienced than explained, which describes most consumer products, should find a way to deliver the experience directly to the decision-maker, without an invitation if necessary.
Notice the craft as well. The order was small, specific, and fun. It was beer, not groceries. It arrived at the office. It needed no explanation beyond "This is Instacart!" A demonstration that requires instructions is just another pitch.
Tactic: When you are rejected on a description, find the smallest, most delightful version of the product and deliver it to the rejecting person unannounced.
Principle 6
Be your own first shopper
Mehta filled the first order himself and tipped himself. He let customers order when no shoppers were online, then dropped everything to fill those orders. His team bought one of every item at Trader Joe's and photographed the whole store. All of this was the opposite of scalable, and all of it was deliberate. "I find that this is one of the biggest competitive advantages that a startup has over a larger company," he said.
The logic is about sequence. Demand has to come before scale, because scaling something nobody wants is just a quicker route to failure. A founder doing the work by hand learns where it actually breaks, which aisle, which shelf, which item is always out of stock, and that knowledge becomes the spec for the software that eventually replaces him. Instacart's catalog, routing, and batching systems were, in a sense, a record of what Mehta and his team had done on foot.
The payoff was measurable: ten cities, 10 percent week-over-week growth for twenty straight weeks. "We realized we had the best product in the market." That was not a guess. They knew because they had been the product.
Tactic: For the first hundred transactions, perform the fulfillment yourself and log every point of friction. That log is your first engineering roadmap.
Principle 7
Read the existential threat as a recruiting event
When Amazon bought Whole Foods, the press saw a bomb dropped on Instacart. Mehta saw it dropped on the whole industry, and he was right. Every grocer in North America suddenly faced a competitor with Amazon's capital and logistics, and none of them had a neutral way to sell online. Instacart did. The threat that seemed fatal for Instacart turned out to be fatal only for the grocers' indifference to Instacart.
T
Reframing the Whole Foods Shock
The same event, read two ways.
| Conventional reading | Mehta's reading | Outcome |
|---|---|---|
| Lost our biggest partner to our biggest rival. | Every rival grocer now needs us. | Signed every major North American retailer |
| Amazon now owns grocery. | Amazon now threatens grocery. | Costco, Kroger align with Instacart |
| Retrench and defend. | Expand and sign. | Canada via Loblaw; Unata acquisition |
| Whole Foods relationship is over. | Relationships are contingent. | Whole Foods Canada returns, 2024 |
The underlying principle: when a dominant player makes a move that hurts you, ask who else it hurts more, and whether you are the obvious ally for all of them. Neutral intermediaries become far more valuable the moment one participant turns aggressive.
Tactic: When a giant makes a move against you, list every other company the move threatens, then call each of them that week.
Principle 8
Find the weightless cargo
In 2013 Mehta mocked AmazonFresh for loading Xboxes onto grocery trucks to subsidize low-margin food. By 2018 he had found his own subsidy, and it had no weight at all: advertising. Thousands of consumer brands were paying to promote their products to shoppers who were already in the virtual aisle. The four-sided marketplace (customers, retailers, shoppers, advertisers) turned a thin-margin delivery business into a media business with a delivery business attached.
This is the lesson Webvan never got to learn. In thin-margin categories, the margin is usually not in the core transaction. It is in the attention the transaction gathers. Grocers had always charged brands for shelf placement, and Instacart rebuilt that economics in software. Bloomberg later credited the "pivot to a higher-margin advertising business" for the company's IPO-era revenue growth, and the post-Mehta Instacart Platform made the model explicit.
The Xbox critique was correct about the cargo and wrong about the principle. The bundle was right. The item in it was wrong.
Tactic: Identify what your core transaction makes visible, including attention, data, and placement, and ask who would pay to be seen there.
Principle 9
Put everyone on the shop floor
At Instacart, every corporate employee worked as an in-store shopper during their first week. Mehta said it kept staff "connected to our customers" and taught engineers the product fast. It was also a way to keep his own first act alive in the company: the founder who filled his own order, made into policy.
The deeper value is epistemic. In a marketplace, the people writing the software rarely do the work the software directs. An engineer who has stood in an aisle looking for an item the app insists is on shelf six writes better inventory code. A product manager who has carried bags up three flights of stairs thinks harder about batching. The practice cannot erase the distance between headquarters and the shop floor, and Principle 10 shows what remained, but it narrows it.
Tactic: Make every new hire do the frontline job of your marketplace's least-visible participant before they write a line of code or a strategy memo.
Principle 10
Price the human layer before someone prices it for you
This principle is a warning drawn from Instacart's record, not a celebration. The asset-light model depended on hundreds of thousands of contractors, and for a decade the company repriced their work one disputed change at a time: the optional service amount customers mistook for a tip, the mandatory 5 percent fee, the default tip raised from zero to five, the pay-system change of late 2018. Each change brought backlash, and the backlash came from outside the company: a class action settled for $4.6 million, a union election in Skokie, a walk-off, letters from two U.S. senators, a sick-leave agreement with the D.C. attorney general. After Mehta left, an FTC settlement for $60 million continued the pattern on the customer side.
In each case the company ended up making changes it could have designed in advance, on worse terms and under public pressure. The pandemic shows this most clearly. Instacart called shoppers "household heroes" and added hundreds of thousands of them, while its sick-pay policy drew criticism as too hard to qualify for and its safety kits arrived late.
The lesson for any marketplace founder is that the fee, the tip, and the worker's classification are not back-office details. They are the product, as seen by the people who keep the platform running. In the first order, Mehta paid himself generously because he knew exactly what the trip had cost. Every subsequent dispute was, at bottom, about whether the company still knew.
Tactic: Write down, in plain language, what a frontline worker on your platform earns per hour after expenses. If you would not defend that number to a senator, change it before one asks.
Principle 11
Hire your successor while you still can
In 2017 Mehta said that "your company tends to reflect your personality," and that a founder's job is to know his weaknesses and hire people who are better at them. In 2021 he applied that to the top job. He brought in Fidji Simo, who had helped Facebook scale from 1,000 to nearly 100,000 people and had lived through its move from private to public. That was exactly the next phase Instacart faced, and Mehta had not done it before.
The surrounding facts were messier: a reported DoorDash sale attempt, press describing chaotic product management, a valuation about to be cut three times. That messiness is part of the lesson. Founders who wait until the board forces the question lose control of the answer. Mehta, whatever the pressures, chose his successor himself, described the choice in his own words, and kept the line that he could have stayed if he had wanted to. Under Simo the company went public, turned profitable, and grew its platform business.
Tactic: At each order-of-magnitude change in headcount or the start of IPO preparation, ask honestly whether you are still the best person to run the next phase. Answer before the board does.
Principle 12
When you leave, leave
Many founders become executive chairmen and stay that way for years, half in and half out, making the new CEO's job harder. Mehta tried the role for a year and decided it was the wrong shape for him. "I want to pursue a new mission," he wrote, "and I want to do it with the same singular focus that I had while building Instacart." He left the board on listing day, sold a modest slice of his shares, and was already running Cloud Health Systems.
The phrase singular focus carries the principle. Mehta had used it in 2019 to describe his commitment to Instacart. In 2022 he used it to justify leaving. A founder whose main skill is total attention cannot split that attention without weakening both efforts. A clean exit is good for the successor and also preserves the founder's own capacity to start again.
Tactic: If your role at a company you founded no longer requires your full attention, set a fixed date to leave it entirely, and announce that date publicly.
Free playbook
Get The Business Model Playbook
58 business models, one visual page each: how the money flows, the metrics that matter, and who runs it. Free when you join the Faster Than Normal email.
Free. No spam. Unsubscribe anytime.
Part IIIQuotes and Maxims
In their words
— Apoorva Mehta, Startup School NY, 2014As a founder, you have to be extremely resilient. You have to go from failure to failure without losing any steam.
— Apoorva Mehta, Startup School NY, 2014The reason to start a company should never be to start a company. The reason to start a company should be to solve a problem that you truly truly care about.
— Apoorva Mehta, to Fortune, 2017From Steve Jobs and Jeff Bezos, the biggest thing that I have learned is to not to conform. In fact one of our company values at Instacart is called 'Of course, but maybe.'
— Apoorva Mehta, to Fortune's Susie Gharib, 2017It is very important for you to understand who you are, what your strengths are, what your weaknesses are, and what your motivations are. I have found that a lot of times people end up deceiving themselves.
— Apoorva Mehta, to Fortune, 2015Taking one mandatory day off every week has done wonders for my productivity. I realized a few years back that I used to burn out every few months, and it was happening increasingly.
Maxims
- Care first, then build. Twenty failures taught Mehta that a product you don't think about at night is a product you will not make work.
- Look in your own refrigerator. The best founding problems come from a friction you have personally carried, preferably for years.
- A famous failure is a dated document. Ask what has changed since the floppy disk was written.
- Route, don't own. Use existing stores, cars, and shelves, and write the software that connects them.
- Deliver the product to the person saying no. Gatekeepers reject descriptions far more easily than experiences.
- Do it by hand until it hurts. Buying one of every item is how you learn what the catalog should be.
- A rival's power grab recruits your allies. When the giant moves, call everyone it threatens.
- The margin is in the attention. In thin-margin categories, the most valuable cargo often weighs nothing.
- The tip is the product. How a marketplace pays the person carrying the bag is a design decision, not an afterthought.
- Leave on purpose. Choose your successor, then give the next mission your full attention.
