This article appears in VICE Magazine's Algorithms issue, which investigates the rules that govern our society, and what happens when they're broken.
In 2019, Lisa, a single mother, lost her job as the media relations director at a firm in Nashville, Tennessee, moved south to her hometown Jacksonville, Florida, and signed up to deliver groceries for Instacart.
With the $1,000 she earned each week shuttling groceries from supermarkets to her clients, she could make payments on her $2,300 a month mortgage, and car and utility bills, with enough left over to order novels from Amazon and household gadgets from QVC.
The arrival of COVID-19 pandemic in Florida changed all of that. In May, Lisa’s weekly earnings hurtled downward from $1000 to $500 to $90 (recovering partially in June and July). She spent her days glued to her phone, refreshing the Instacart app for hours. Each time a new batch, or set of delivery orders, would flash on her phone screen, someone else in her area would snatch them up in a matter of seconds, often before she had time to react.
The pandemic had prompted more households than ever before to sign up for Instacart’s grocery delivery service, making it profitable for the first time since its founding in 2012, but Lisa—a pseudonym, so she could speak candidly about her working conditions—was barely scraping by, and had picked up a second part time job at a local Publix grocery store to make ends meet.
“I had to be refreshing my phone all the time, sometimes waiting for three hours that I wasn’t getting paid [for], before I got a good order,” she told Motherboard. “You have to be attached to your phone. I couldn’t do anything else.”
In late April, the San Francisco-based start-up announced it had hired 300,000 new gig workers to meet rising demand for grocery delivery during the first two months of the pandemic, and had plans to bring on an additional 200,000 workers. The hiring boom in on-demand grocery delivery across the country wasn't limited to Instacart. The Target-owned delivery platform, Shipt, hired tens of thousands of shoppers during the pandemic. As of early August, the delivery service Amazon Flex was actively recruiting in 17 cities. Almost overnight, hundreds of thousands of laid off or underemployed Americans had transformed into grocery delivery gig workers.
But the demand for groceries didn’t keep pace with the number of Americans flocking to grocery and food delivery apps for available work, and workers found themselves competing against each other while refreshing their phones into oblivion. If heartwarming stories of mutual aid, where communities have banded together to care for, shelter, and feed one another during the pandemic have shown the power of solidarity in the absence of a strong social safety net, then the hundreds of thousands of desperate Americans gravitating towards gig work, and competing under and against a regularly shifting algorithm for an elusive supply of grocery delivery orders, paint a less inspiring picture.
A spokesperson for Instacart told Motherboard that shopper earnings have actually increased from their pre-pandemic levels, based on customer and shopper marketplace data. "In the wake of COVID-19, shopper earnings have increased by as much as 60% and shopper NPS - a measure of shopper happiness and overall sentiment - is at the highest level in company history," the spokesperson said.
"To further support this community, our team has introduced new recognition programs, product features, shopper perks, and resources to enhance the shopper experience," they continued. "We’ll continue to invest in this important community as we focus on delivering the best possible experience for all shoppers across North America.”
In early 2020, Lisa ranked near the top of 390 Instacart shoppers in the Jacksonville Beaches metro area, a sprawling island community to the north of Florida’s largest city. But during the pandemic, she says the app had hired an additional 900 shoppers in her area.
Adding to her difficulties, in March, Instacart had suspended its algorithmic ranking system, which rewards gig workers with the highest customer ratings by offering them the most lucrative orders first. The company explained in an announcement that, due to the “national state of emergency, all ratings below 5 stars” would be “forgiven.” (As of July, Instacart had reinstated the ranking system.) For veteran Instacart workers like Lisa, who had worked hard to achieve near-perfect ratings over hundreds of orders, this meant drastically lower earnings. As another Instacart shopper explained to me, the difference between a perfect 5 star average rating and a 4.97 star average rating can mean everything for access to the most lucrative orders under normal circumstances.
At the same time, in Instacart forums on Facebook, rumors circulated widely that the decline of work was primarily linked to the rise of third-party bots which snatched up orders quicker than a human could, and sold them to shoppers willing to pay a high price to game the system.
“When coronavirus got bad here in May, third-party bots popped up. We ran into them at stores. You could tell who they were because they didn’t know the layout of the stores. One guy bragged about it to me in an Aldi grocery store. He was a bot,” Lisa said. (Shoppers often refer to other shoppers who pay for automated tools as “bots.”)
Some of the rumors were blatantly racist, accusing undocumented shoppers of using the bots to game the system. Motherboard viewed posts on Instacart Facebook groups that linked the bots to rings of non-English-speaking, Latinx immigrants who shopped in large groups.
Multiple opportunists had developed bots that gave gig workers who paid, in some cases, thousands of dollars, the advantage of being able to scoop up orders faster than those who didn’t. But skeptical Instacart shoppers say the bots were mostly scams, the rumors were racist, and bots were not responsible for a significant portion of competition for orders—Instacart’s massive hiring spree was.
“Bot services have existed for Instacart for nearly two years, but bots have never been nearly as popular or common as many shoppers believe they are,” Heidi Carrico, an Instacart shopper and organizer with Gig Workers Collective, said. “The true problem is that Instacart hired 500,000 new Shoppers, while demand and order volume has declined from its peak at the beginning of the pandemic, creating a dynamic of fierce competition for orders.”
Instacart has claimed that the bots violate the companies’ trademarks and terms of service, and sent cease-and-desist letters to third-parties. In late July, Instacart sent out an email informing shoppers that it has partnered with a security platform “to develop a bot bounty program, specifically built to combat bots on the Instacart platform,” and promised to deactivate any shoppers who used third-party apps.
Veteran Instacart shoppers have noted that the pandemic is not the first time they’ve dealt with saturated markets and times when it’s difficult to get orders. Most years, usually in the fall, for reasons not entirely clear, orders slow down, competition rises, and Instacart slashes pay, prompting some contingent of shoppers to quit working on the app.
“Things have started to get better,” Lisa told Motherboard in July. “It’s a combo effect. The bots are being handled and slowly but surely the oversaturation is starting to come down. All of these people who signed up to delivery groceries thought it was easy. Now we’re back to our rating system and that’s how we get batches.”
Though gig workers that Motherboard spoke to said their earnings recovered partially in June and July, suggesting that some of the new hires quit working on Instacart once restaurants and retail stores reopened, longtime workers say the number of shoppers are still above pre-pandemic levels and wages remain down.
Bill, an Instacart shopper in Houston, originally started working on the app nearly two years ago to supplement his income (he previously owned a food truck), and was finding it hard to quit gig work, until the pandemic hit and helped him with the decision.
“I was making $700 to $800 a week, then things just went bonkers. The orders were coming in so fast, you couldn’t swipe fast enough. I don’t know how a human can react that fast. Suddenly I was making $60 a week,” Bill said. “I was a mess. It used to be like anything else: ‘Doing this sucks but I know I can get batches, and I know I can make at least $500.’ All of the sudden it just disappears. Your income just goes. It’s very stressful.”
While many gig workers who provide services that involve close human interaction (such as Uber and Lyft drivers, TaskRabbit workers who enter client’s homes, and Wag and Rover dog walkers) have seen their income plunge, as the New York Times recently reported, on-demand grocery delivery and food-delivery apps have boomed, making record profits. Experts say that though the influx of grocery delivery workers will likely fall as businesses reopen, many workers who sign up for gig economy apps as a temporary job end up staying much longer than they had planned, either because they can’t find more lucrative job opportunities or because making income on the app consumes so much time and leaves few hours of the day for job hunting.
“You notice people of all walks of life are out of work doing Instacart,” said Bill. “People who work in hospitality, restaurants, bars, and I don’t begrudge them. Everybody’s just trying to do the best for themselves. What sucks most is that it’s taken any hope of making money.”
During past economic crises laid off and unemployed workers have always gravitated to less secure forms of work—hustling to string jobs together. In fact, the emergence of platform gig economy companies grew out of the 2008 recession, a period of high unemployment and slow job growth. Gig economy start-ups Uber, Instacart, TaskRabbit, Lyft, Postmates, DoorDash, and Caviar were all founded in the span of four years, between 2008 and 2012.
“Gig work is often marketed as a solution to unemployment, or economic downturn. A lot of these gig economy jobs have low barriers to entry,” said Alexandrea Ravenelle, a professor of Sociology at the University of North Carolina at Chapel Hill who is leading a National Sciences Foundation-funded study on the impact of COVID-19 on gig workers. “During COVID-19, we’ve seen an influx of people turning to gig work as a social safety occupation of last resort.”
The companies marketed themselves as technological innovations that created millions of new jobs and allowed workers the opportunity and flexibility to pick their own hours and schedules free from the scrutiny of a boss. But the catch was that gig workers, as independent contractors, did not receive any of the typical benefits such as minimum wage guarantees, health insurance, overtime pay, and paid sick leave offered to employees.
Thus, critics of Uber and Instacart charge that their main contribution has been the creation of precarious, low-wage, dead-end jobs that offer no guarantee of job security or income—an unfortunate reality Instacart workers who’ve seen their earnings diminish overnight learned during the pandemic.
Marvin, a shopper who lives outside Huntsville, Alabama, took up Instacart in April 2019, for supplemental income. When he lost his job as a flight attendant last fall, it became a full time gig.
Marvin also saw his wages steadily decline during the pandemic. He was forced to dip into his savings, often spending 10 hours-day waiting in parking lots “listening to every song in the world” and refreshing his phone until orders came in. “No, I don’t enjoy sitting in my car for that long,” he told Motherboard. “I survived on a lot of Spam.”
At the same time, the number of Instacart shoppers in his metro area tripled from 40 to 120. (We know this because Instacart posts the number of active shoppers in each metro area on its app by rank.) “A lot of waitresses, waiters, teachers, housewives, and exotic dancers came out of the woodwork,” said Marvin, who has befriended many shoppers in the area.
Marvin, who is 28, says he cuts out the unpaid time he spends waiting by juggling Instacart work with less lucrative orders on DoorDash in his area, usually at fast food restaurants, which often pay less than $10. The goal is never to earn less than he spends on gas and the costs of maintaining his car.
On a day in late July, Marvin logged his order for Motherboard; he visited Kroger three times with trips interspersed to McDonald’s, Arby’s, and Chick-Fil-A, earning $99.33 in six hours, and sometimes traveling up to 17 miles to make a delivery.
“I try to fit in one or two DoorDash orders for every Instacart order. It doesn’t always work out that way but that’s what I try to do not to have any idle time,” Marvin said. “I’ll often pick up $5 orders from Taco Bell or McDonalds.”
For Instacart and other gig workers, learning to identify lucrative orders takes time and practice. Workers told Motherboard they tend to gravitate to the stores with layouts they know best, customers who’ve tipped and rated well in the past.
“Our ranking [and ability to get good orders] depends on what the customer rates. It’s solely up to them,” Lisa, the Instacart shopper in Florida, said. “We’re at their mercy. If they don’t like how something was bagged, they could give us a four. I had one lady say, 'I never give out a five star rating.' I don’t want to be dinged for that.”
Workers prefer orders with shorter distances, tending to avoid apartment buildings with long flights of stairs. Similarly, accepting orders with multiple repeat items from different customers like frozen pizzas, cat food, or bananas is easier than traversing back and forth in search of obscure products.
“At first it was taking me four minutes per item at my local Sprouts supermarket. Now it takes 30 seconds,” said Bill, the Instacart shopper in Houston. “You learn quickly that you make more money the faster you can complete an order.”
Now that Instacart has reinstated its ranking system, veteran shoppers say they hope to see their earnings increase, as they’re awarded more lucrative orders, while shoppers hired during the pandemic have to undergo the learning curve that they once did. Experts say that it’s difficult to know at this point how the surge of workers competing for gigs on food and grocery delivery apps will last; due to poor reporting systems, there aren’t reliable statistics on gig workers.
“The Great Recession brought about an economic contraction and lots of people took up gig work. It was hard for people to get back into jobs they once were in,” Ravenelle, the sociologist at UNC Chapel Hill said. “We’ll probably see something similar in terms of COVID. We’ll feel this for a very long time.”
Follow Lauren Gurley on Twitter.