Agile Ethics Rideshare Roleplay
https://miro.com/app/board/uXjVIYcaTAY=/?share_link_id=295128990174
Sprint 1 – Tiger Hunts and Quests
You first try “Tiger Hunts.” To push a pack of drivers to a spot with a lot of riders, you temporarily increased the fares for that area, so even short rides were worth it. At first it worked. Lots of drivers would show up; every rider would get picked up. But riders complained that “Hunts” drove the price up when they _really_ needed a ride. And drivers complained that if they arrived a minute too late, the fares would already have dropped. They felt used. Experienced drivers told younger ones, “Don’t Hunt; you’re the prey.”
On the upside, now you have data on where and when demand usually spikes. Your product owner suggests a feature called “Tiger Quests,” a gamified “challenge” system, where some drivers can win prizes for working in certain pre-announced areas/times. You also have data about drivers’ habits, so you can use a machine learning algorithm to target precisely those drivers who would not otherwise drive in those areas/times, and offer them incentives (meaning you don’t waste money on drivers who’d be there anyway, or on incentives too large). Since Stryped is expanding, you also want to make the experience rewarding for new drivers – if you see them accepting challenges and winning prizes, that’s proof “Quests” are working.
Quests could be a win-win – more transparency on fares and more driver choice. You plan to advertise it to drivers, saying “Fares as clear as a tiger’s eye.” But others in the company think it’s kind of creepy to control drivers, like you’re a big cat toying with mice.
Should you stick with “Hunts,” out of fear of the effects of personalized games? Or roll out “Quests” to balance supply and demand?
Stick With Hunts/Roll Out Quests
Sprint 2 – Tiger Swipe
Your CEO chooses to deploy Quests, despite some reservations from the team. And at first, drivers are enthusiastic about the gamification, and feel like they’re more in control of their fate.
But soon the old sense returns that they’re being used. The algorithm was personalized to their driving habits, so it optimized by treating them differently. When they live in a high demand area, others are given prizes to go there but they are not, leading to different pay for similar rides. New drivers get lots of Quests, but experienced drivers do not. Drivers start to share stories of the algorithm sending them far away from busy areas when they were _just_ about to win a prize.
Your user experience designer (who used to work at Tinder) proposes a feature called “Tiger Swipe.” Just like a dating app, Swipe would show multiple possible rides to a driver, with the location, fare and customer info. If they’re willing to take that ride at the offered fare, they swipe right. New, old, no matter where they live, they get the chance to see fares and decide what’s right for them. You would publicize it to them, saying “Choose your stripes with Tiger Swipe.”
But your chief economist is reticent. Swipe would show the same ride to many drivers at different fares, and give it to the one who accepted the lowest fare. But she’s worried that if you optimize the algo for profit margin, it will start offering lower and lower fares.
Should you stick with “Quests,” and accept its different treatment of drivers? Or introduce “Swipe,” to let all drivers choose if a fare works for them?
Stick with Quests/Introduce Swipe
Sprint 3 - NYC Makes You “Pay Up”
Tiger Swipe backfires. You had wanted to increase driver autonomy without lowering wages. But, drivers reported that Swipe gave them anxiety: they felt compelled to swipe on an offer right away or risk losing the fare. It was like gambling: even if the bid was too low they’d swipe right on it, to get the reward. (Sometimes while driving. Unsafe!)
Worse, drivers say their pay is unpredictable, and dropping. Some troublemakers at the University are gathering data from your drivers on an app they call FairFare. They claim your “take rate” is increasing – the driver’s cut is down, while you make more. And journalists show that more than 60% of the time different drivers are offered different rates for the same rides.
Faced with all this, NYC passes a minimum wage law for gig drivers. If a driver doesn’t make a minimum in fares per hour, Stryped will have to “pay up.” But the City is wise to your manipulative ways. They know you could push lots of drivers to be out when you want prices low, then turn around and get them to drive in the dead of night and in odd places where you can charge more and pay more. At some times drivers would be cruising the city in an empty car (wasting gas they pay for). And they’d be dead tired from odd hours. So the City law says that the lower your drivers’ average “utilization rate” (time with a rider in the car), the higher the minimum pay will be.
Stryped is shocked! The City’s new plan will incentivize drivers to be out when there are _few_ riders. To stop that, you could do one of two things:
You could limit your number of drivers by halting new signups. But the fewer drivers, the higher the price and the longer the wait time for riders. Or you could simply limit the number of drivers at any given time by doing temporary “lockouts.” The app would randomly shut off for some when there aren’t a lot of riders, only to blink on later. Since it’s not targeted, it’s fair.
Should you try “lockouts” and keep Stryped open to a lot of drivers? Or limit new signups, giving fewer the opportunity to have a good side-hustle?
Try Lockouts/Limit Drivers
Sprint 4 – Battle Over Deactivations
Low prices and “gig” work are in your DNA. So your CEO decides on “lockouts.” But there’s an outcry from drivers since it seems so random (it is!). No one wants to lose the lockout lottery when they have kids to feed.
So now you have to limit numbers. You start by “deactivating” drivers you claim are bad, rude and dangerous, with low star ratings. But drivers claims they’re actually getting deactivated for driving at hours you don’t want, or for rejecting a lot of rides to choose only the best fares.
So the City passes a new law that Stryped must: 1) Do immediate deactivations only for cases of “egregious misconduct,” like harassment and robbery. 2) Base any other deactivations on clearly stated policies that do not include hours of work. And 3) Provide drivers 14-day notice of a planned deactivation, and a human review of their case.
Some at Stryped say you can fight this new law. PR will tell the public it would lead to unsafe drivers at their houses, and lost privacy when their driver complaints are made public. Legal says they can convince a judge that your freedom of association (in the U.S. 1st Amendment) will be lost if you can’t choose to do business with whomever you want. And they will argue that you ought to have privacy of information (4th Amendment) in the data you use to create a profitable business.
Should you file this lawsuit, and start a PR blitz? Or should you change your Strypes, becoming more like a taxi company by scheduling driver hours and paying minimum wage?
Lawsuit, PR Blitz/Change Your Strypes
______________________________________
Sprint Zero
You take the City to court, but lose. So Stryped can no longer do deactivations. You decide to charge a clearly-labeled $5 service fee to make up for lost profits, but ridership falls by 45%. Stryped tries to raise another round of capital to scale – maybe your lower profit per ride can be offset by getting more rides in cities all over the country. But the venture capitalists don’t bite.
Your CEO issues a press release saying that the City’s law “definitely hurt the people that it’s supposed to protect.” Some drivers disagree, saying they’re making more across the board. But most say they’re not – each ride pays more, but there are so many fewer rides.
The city relents, reducing the wage required, and making it a bit harder for deactivated drivers to sue you.
You’re still in business. But at what cost?
______________
References:
- Rideshare Transparency: Translating Gig Worker Insights on AI Platform Design to Policy. CSCW 2025
- Policy memo
- FairFare - paper under review (attached PDF)
- Trip Radar: Uber drivers warn new feature is ‘putting you at risk’
- Bloomberg story on NYC Min Wage Laws: NYC Lockouts
- Algorithmic Wage Discrimination - Paper, More Perfect Union Story
- Deactivation with and without Representation: The Role of Dispute Arbitration for Seattle Rideshare Drivers
- Uber CEO says minimum wage law in Seattle has ‘hurt the people that it’s supposed to protect’
- More Perfect Union Story on the WAO: How Uber Robs Its Drivers: Uber and Lyft are hiding the true cost of a ride for a reason
Re: Lockouts
https://www.nytimes.com/2024/12/13/nyregion/uber-lyft-lockouts-wages-nyc.html
Re: Deactivations
https://www.seattle.gov/council/issues/past-issues/payup
https://www.seattletimes.com/seattle-news/politics/uber-sues-seattle-over-driver-deactivation-law/
Re: Minnesota Law