Dozens of us crowd around the coffee shop’s tables, eyes glued to our phones, waiting for something — anything — to come through. Our cube-shaped bags clutter the floor of the Tim Hortons at Dundas St. W. and Spadina Ave., a popular waiting spot for food delivery couriers.
An occasional “ping” cuts through the idle chatter, jolting us to attention.
I snatch my phone, hoping for a notification — but my screen is blank. It’s another courier’s order and he’s already bolting out the door, bag in hand.
It’s my first week working as an Uber Eats courier and despite it being peak time for deliveries on a Thursday afternoon, I haven’t received an order in over three hours.
I turn to the driver next to me. Is my app malfunctioning or am I doing something wrong? After all, the app’s city map shows I’m in a “hot spot,” a high-demand area during high-demand hours.
He shrugs. He spent $10 on his GO Train commute to the city, he tells me, and has only made $8 on Uber Eats all day.
“Too many drivers and not enough orders,” he says.
Toronto’s sidewalks and congested streets have never felt more crowded with food couriers. Spurred by a pandemic surge in delivery app usage, the food courier workforce has ballooned, increasingly made up of newcomers and precarious workers for whom few other job options exist.
In 2023 alone, Statistics Canada says the workforce for delivery apps like Uber Eats grew by a staggering 46 per cent.
In taking a job as a delivery worker for Uber Eats, the city’s most popular food delivery app, I joined the ranks of an oversaturated workforce, where on any given night, a surplus of food couriers outnumbers the available orders.
Behind the boom lies a troubling trend: couriers’ pay and behaviour are governed by opaque algorithms that determine wages based on hidden criteria. Using artificial intelligence technology, these platforms keep drivers tethered to the app, waiting unpaid for their next order.
For drivers, the results are unpredictable and too often unfair. Data obtained by the Star shows Uber Eats’ platform can offer two food couriers different wages for the exact same trip.
Labour advocates charge that the app collects data on driver behaviour and can use it to decide who it can pay at a lower rate, allowing the company to pocket the difference and boost its revenue. This concept is widely referred to as algorithmic wage discrimination.
“The app has total control over how a worker gets paid,” says Veena Dubal, a University of California law professor whose research focuses on the gig economy.
“Minimum wage and the idea that hard work should lead to economic security, can be — and are being — destroyed by these A.I. systems.”
Uber says Star reporter’s experience ‘atypical’
The Ontario government’s legislation aimed at improving conditions for digital platform workers won’t take effect until 2025 — and even then, workers say it will fail to address the most pernicious elements of gig work.
Uber says my experience as a food courier “was atypical” and didn’t match that of an average Toronto delivery driver.
The company denies allegations of algorithmic wage discrimination, stating that it does not use a driver’s “past behaviours” to determine who it can pay at a lower rate.
Uber says it’s transparent with drivers about potential earnings. The company says delivery workers choose the platform for its flexibility and their ability to “earn money on their own terms.”
I worked as an Uber Eats courier for six weeks. I wanted to understand the costs behind the convenience of app-based delivery and gain insight into how gig work is evolving as algorithms call the shots.
Many of the food couriers I worked alongside were young international students struggling to earn an income while they make their way through school. Others were refugees or undocumented workers, navigating precarious lives.
Of the dozens of workers I spoke with, almost none of them have permanent status in Canada. Most were afraid to speak openly about their experiences with Uber Eats or other apps they deliver for out of fear of jeopardizing their livelihoods.
I rented an electric bike and set out to complete at least 100 orders. The goal was to collect independent data that could shed light on how drivers’ wages are determined. But I soon discovered I’d get far fewer orders than expected, a struggle many couriers say is becoming all too common.
After my first 20 hours on the Uber Eats app as a food courier, I made just $28.98.
‘Gamification’ of gig work
In my first weeks delivering food there were hours and even days when I wouldn’t get a single order.
Hot spots on the app could change at any moment. When I followed the prompts and biked to one of these locations, like in Liberty Village, it vanished the moment I arrived. A new hot spot appeared on the other end of the downtown core at Yonge and Dundas, nudging me to wait on sidewalks crowded with couriers competing for the same jobs.
The more time I spent on the app, the more I felt like a player in a game where I couldn’t figure out the rules. But every courier I met had their own theory on how to beat the game.
Some swear by logging in and out of the app to trigger more orders. Others insist on keeping the app open at all times without interruption. Many advise ignoring the heatmap altogether and heading to quieter areas with less competition. Some say to keep biking around instead of waiting in one spot for too long — “the app will think you’re taking a break,” they say.
One thing became clear: chasing orders was part of the job.
“This is the gamification of work,” said Jim Stanford, economist and director of the think tank Centre for Future Work.
Employing hundreds of social scientists and data scientists, Uber has talked about how the company has experimented with video game techniques and other tactics to incentivize drivers to stay on the road for more hours.
However, Uber told the Star this characterization was “misleading,” and that its data science team was “focused on making offers more transparent and improving the Uber Eats app.”
As soon as I logged into the app, I was met with flashing heat maps signalling high demand areas, performance goals tied to rewards programs and countdown timers pressuring me to accept rides while I was in the middle of biking in rush hour traffic.
I noticed that I was more likely to get orders if I was constantly on the move. To test this theory, I logged on to the app and stayed in a hot spot for seven hours. Not a single order came through.
When asked about this experience, Uber Eats said it “does not require a delivery person to be moving to receive trips.”
I changed my strategy and biked in circles around the city during peak lunch and dinner hours. More orders came in this way than when I’d stop and wait at hot spots.
140 hours on the clock — but few of those count as working
Over six weeks, I spent 140 hours and 22 minutes on the app in search of work. But I was paid only for 15 hours and 49 minutes — the time Uber Eats determined I was actively delivering orders.
One courier I met on the road was a 28-year-old from India who had been delivering food for over a year trying to pay off his MBA tuition. Unable to find a steady job, he worked on multiple delivery apps, but said it wasn’t uncommon to go six hours between receiving orders.
Sometimes, he earned less than $10 over an eight-hour workday.
“I’m OK to do hard work, but I’m not making any money,” said the driver.
Like most of the couriers I spoke to, he asked not to be quoted by name for fear that he will face repercussions from delivery platforms or jeopardize his chances of finding better opportunities in the future.
Having work appear scarce creates a sense of competition that makes drivers less selective of what orders they take, said labour relations lawyer Ryan White of Cavalluzzo LLP.
“This feeling of scarcity might lead delivery workers to think, ‘If I don’t take this assignment, even if it’s not an assignment I want … someone else is going to,’ ” White said.
Uber spokesperson Keerthana Rang said it’s “difficult to comment on experiences a specific delivery person has because many factors play a role in earnings, like time of day, demand in the area, weather.”
Delivery workers “control where and when they work, they are free to use other apps at the same time as they’re on Uber, choose which deliveries to accept, and they can deliver using their own vehicle, by bike, or on foot.”
I felt anything but in control.
As I spent more time on the app, another change occurred. Instead of simply allowing me to “accept” an order immediately, the app started prompting me to request to “match” with it, meaning the app would first determine whether I was the best fit for the delivery before giving me the job outright.
More often than not, I’d lose the match to another driver. The power to choose was not mine.
Uber says “there are a number of factors that determine how a match is made, including the proximity to the merchant and estimated delivery times.”
One driver suggested I rush to the restaurant as soon as I got a “match” request, in a bid to be the closest to the location when the app made its decision — more guessing, more waiting.
Often, after a few match requests were given to other drivers, finally receiving an order I could accept felt like a small victory — one that kept me eager and engaged.
Control for drivers is an illusion — it’s really “all in the hands of the apps,” White said.
“Workers don’t get to determine their terms and conditions of employment and they don’t get to set the prices that they’re working for.”
Identical orders, different wages
On a bitter February evening in 2022, a group of Uber Eats delivery workers gathered at Wychwood Barns Park near St. Clair Ave. W. and Christie St. Soaked in freezing rain, they huddled together to test an algorithm that had become their boss.
They ran a simple experiment: logging into the app simultaneously mere inches apart, they strived to document how pay varied for identical orders.
When one driver received an order, they took a screenshot capturing the pay and details before declining the job. If one of the other drivers was then offered the order, they did the same thing, creating a side-by-side comparison.
The Star analyzed the data of six delivery workers collected by Gig Workers United, a network of app-based couriers advocating for better workers’ rights.
Out of 21 assignments offered to at least two drivers in the group, all but one — 95 per cent — revealed pay discrepancies for the same job.
For one delivery, from a Harvey’s at St. Clair and Bathurst to a location near Casa Loma, the first driver was offered $6.81. The second was offered $6.18 — a nearly 10 per cent pay difference for the same delivery.
In many cases, the pay difference for identical deliveries was less than 10 cents, an amount that may go unnoticed.
But given the thousands of drivers and orders around the world, if Uber were to “skim off a few cents” from each delivery, the difference could result in “millions … of dollars in profit across millions of rides globally,” Dubal said.
Alleged wage discrimination ‘upends’ notion of fair pay
Dubal has spent more than a decade researching the ride-hailing and gig economy and how these platforms affect workers’ rights, livelihoods and legal protections.
Her research, which includes interviews with drivers and analysis of their pay patterns, has found the business model the couriers participate in ultimately relies on an imbalance of power and information.
Artificial intelligence technology determines what an Uber Eats worker gets paid. Platforms like Uber use these complex algorithms to streamline customer service while maximizing profits and also dictating worker pay, speed and behaviour.
It has fostered a system, Dubal says, where workers performing the same task with the same skills under identical conditions can receive different pay.
Dubal’s research documented this alleged algorithmic wage discrimination among rideshare drivers in the San Francisco Bay area, comparing the fares they received to what other drivers got.
The data from the Toronto food couriers’ experiment, she says, adds to a growing body of evidence that “proves that Uber Eats has created a system where workers do not earn the same amount even if they’re doing the exact same thing.”
“This upends the notion that there should be and is an assumption of equal pay for equal work.”
Uber has denied the app uses an individual worker’s past behaviour to set different wages.
“Uber Eats does not send lower-paying upfront offers to a delivery person with a history of accepting them. Anything written in your story otherwise would be false,” Rang said.
The Uber spokesperson said that any variation of earnings between drivers for the same trip “is likely due to the technical limitations of GPS.
“GPS Satellite visibility, location settings, and device orientation are some of the reasons why phones next to each other might have different geolocation results.”
According to Dubal and Stanford, Uber Eats is applying a concept similar to one it popularized: consumer price discrimination, sometimes called dynamic or surge pricing, where customers are charged different prices based on what companies think they’re willing to pay.
Now, critics say it appears it’s being used to set couriers’ wages.
This raises serious concerns, particularly given the vulnerable demographics of many workers, Stanford said.
Among the couriers I spoke with was a refugee from Eritrea unable to work legally in Canada, who said he had no choice but to rely on income from food delivery apps despite earning dismal wages.
“I do this or I die,” he said.
In February, Uber reported its first annual net profit since the company went public in 2019. The company booked a net profit of $1.9 billion in 2023, compared with a loss of $9.1 billion in 2022.
Uber says the company’s “profitability has come from the ever-growing volume of trips, which has resulted in our revenues growing faster than our costs.”
Hours after announcing its first-ever annual profit, Uber’s CEO told investors on a conference call that the company is increasingly focused on offering drivers’ orders and rides based on their “behavioural patterns.”
“I think what we can do better is targeting different trips to different drivers based on their preferences, or based on behavioural patterns that they are showing us,” said CEO Dara Khosrowshahi, saying it will lead to shorter wait times for customers, happier drivers and a more efficient system.
“That is really the focus going forward: Offering the right trip, at the right price to the right driver.”
A change to the Uber Eats payment structure
Toronto was the first city in the world where Uber Eats operated, launching onto the food delivery scene in 2015. Uber was competing against other companies such as DoorDash and Hurrier.
To ensure there were enough couriers available at all times to meet customer demand, delivery platforms offered workers incentives to keep engaging with their apps. Companies offered “boosts,” which multiplied courier earnings if they delivered food in areas seeing a surge in demand, recalls Brice Sopher, who has spent nearly a decade as a bike courier in Toronto.
“Uber paid so well in the beginning that you didn’t even care about tips — it was that good,” said Sopher, who is also the vice-president of Gig Workers United.
When the pandemic hit, the use of online food delivery platforms surged as customers were confined to their homes and restaurants shuttered their doors.
In June 2020, Uber scrapped a payment structure that had offered couriers clear pay based on pickups, drop-offs, distance and time.
It was replaced by algorithmic pricing. Drivers could no longer see how their base pay was being calculated. For example, I have no way of knowing why a 2.5-kilometre delivery earned me $4.06 before tip, while a nearly identical 2.3-kilometre delivery paid $6.08.
Uber maintains the changes made wages more transparent. Before accepting a delivery, drivers can see fares, including estimated tip, trip distance, estimated time to completion and locations for pick-up and drop off. “The delivery person is in control in determining whether they wish to choose or decline the offer,” Uber’s spokesperson said.
‘Platforms take advantage’ of workers’ desperation
The upfront fare and estimated time calculated by the app often failed to account for delays beyond my control, from waiting for food that wasn’t ready at the restaurant to climbing flights of stairs at a condo tower because the elevators were down. When these incidents happened, my deliveries took longer but my pay stayed the same — this meant I would be earning less money per hour.
Uber said it aims to “minimize delivery person wait time by adjusting expected merchant preparation time based on actuals.”
Demographics of the gig-economy workforce have also changed. Data shows that the surge of ride-hailing and delivery gig workers has been driven largely by newcomers.
Landed immigrants accounted for nearly 60 per cent of people who provided either personal transport or delivery services through an app or platform in 2023, according to data from Statistics Canada.
One courier I spoke with, a 19-year-old university student from India, came to Canada in search of a good education and better living standards. Instead, he says he’s barely scraping by, struggling to pay his tuition at Niagara College. These days, he considers himself lucky to make $50 on a weekend shift.
He searched for work for six months. “When I couldn’t find anything I started doing Uber (deliveries) full-time,” he said.
With low barriers to entry, app-based delivery jobs can be appealing to newcomers who are often shut out of other jobs in the labour market, Stanford said.
High levels of immigration “combined with the weakening of the labour market, means there are more people willing to do this lousy work than there were two years ago,” he said.
“They’re desperate and the platforms take advantage of that.”
A shocking wage
Over six weeks, I hustled to complete 56 orders.
I spent more than 140 hours glued to the app and biking around the city in hopes of coaxing an order out of an algorithm, and made just $243.82, plus another $73 and change in tips. This money is being donated to charity.
In the end, I earned a shocking $1.74 per hour online.
Uber says the average delivery person in Toronto is engaged four times more on an hourly basis than I was during my six-week experiment. And the company reports that the median driver earnings in Toronto in late 2023 was $33.35 per “engaged hour.”
Uber says earnings “must be calculated against engaged time” because a delivery person can have multiple food-delivery apps open at once to optimize their earnings.
Even by that standard, my earnings fell short.
I made approximately $15.41 per engaged hour before tips, below Ontario’s $17.20 minimum wage.
The per-hour pay fluctuated on each order. For one delivery, I made the equivalent of $34 per hour. On another, a paltry $6.95 per hour. In fact, on roughly half of the orders I delivered, I earned the equivalent of less than minimum wage.
The effort didn’t even cover my expenses. Renting an e-bike alone cost me $460 for those six weeks.
The Ontario government has brought forward legislation meant to improve working conditions for gig workers such as food couriers and ride share drivers.
The Digital Platform Workers’ Rights Act will require companies to provide workers with a description of how their wages are calculated, and will guarantee gig workers earn minimum wage — but only for engaged hours.
But the legislation, introduced in 2022, does not come into effect until summer 2025.
“It is important that the government get this right, which includes consulting with workers, digital platforms, the legal sector, and other affected stakeholders,” a spokesperson for the Ministry of Labour said.
Calls for reform for gig work
Some cities have taken steps to regulate platforms such as Uber.
Earlier this month, Toronto city staff recommended capping the number of rideshare licences in a bid to address the ride-hailing industry’s negative impacts on traffic, emissions, and public transit. A staff report concluded that drivers’ median income, when accounting for all the time spent on the app, was just $5.97 an hour after costs such as fuel and insurance — a figure Uber disputes.
The city’s executive committee voted to send the proposal back to the drawing board.
Other cities, such as New York City, have focused on guaranteed wages for drivers.
By paying drivers a guaranteed wage, Stanford said the hope is that this will dissuade companies from courting a surplus of workers to compete for orders.
Critics say that delivery apps for years have gotten around paying drivers minimum wage by misclassifying gig workers as independent contractors — excluding workers from the full set of rights they would otherwise be entitled to as employees.
Organizations like Gig Workers United and RideFair TO have been advocating for gig workers to be classified as employees, which would grant them access to essential benefits, including employment insurance and minimum sick days.
App companies have pushed back saying this would undermine couriers’ flexibility to work across multiple platforms and to set their own schedules.
Sopher, who has worked on more than one app at once, described the experience as “degrading.”
“It’s like working two or more jobs at the exact same time and still not making enough money,” he said. “You only do it because you have to.”
In Uber’s 2023 annual report, the company said its business “would be adversely affected if drivers were classified as employees, workers or quasi-employees instead of independent contractors.”
Dubal warns that the ripple effects of this algorithm-driven system, in which drivers have minimal protections, extend far beyond food delivery.
This same technology is already playing a significant role in managing jobs in some distribution warehouses, from scheduling shifts to assigning tasks and even reportedly firing employees.
For six weeks, I knew I had the option to log off whenever I wanted, to step away from the relentless chase for orders. But for a growing number of workers, their primary — often sole — source of income is determined by an app, with wage rates and assignments beyond their control.
Sopher, who has all but given up on delivery work, fears that if companies like Uber continue to misclassify delivery workers “you’re going to have a permanent two-tier employment regime in Ontario.”
“It’s all of the worst parts of being an employee and all the worst parts of being an independent contractor rolled up in one.”
