Evidence of algorithmic management at PickMe
| Representative data | First-order concepts | Aggregate dimensions |
|---|---|---|
| “But obviously, I mean, we know how many hours they put, right? If we know that he’s putting twelve hours or eight hours… if it was eight hours, we know that he’s working full-time. But if he ends up putting four hours, for someone who’s, then we know there’s something going on, right? So, we just need to know how to manage that situation” (SM1) | Using information about working hours to penalize them for using other ride-sharing apps | Data and surveillance |
| “We monitor his behaviours, performance through our application, so we have built some algorithm to make sure that we capture that information. So, we periodically review those” (SM3) | Monitor and aggregate data on driver behavior for review and response | |
| “So, we’ve been monitoring the rating continuously” (SM3) | Continuous monitoring of rating | |
| “So, we have different loyalty years from a newbie to platinum, so that depending on where they land in the listing, depending on how long they have service, what is their rating, how many rides, jobs they have done…We will decide their rating, which is every month, it’ll get updated” (SM3) | Gather various information on the app to determine rating | |
| “We do a lot of changes to our technology to give a better user experience” (SM3) | Adapt technology to respond to driver | Responsiveness |
| “We give periodical feedback to the driver also” (SM3) | Feedback given to driver | |
| “We are running some machine learning and different algorithms now for handle… we have a kind of a weighted average. If let’s say the driver also can rate the customer, we look at the person. To create a passenger rating too for a driver, the weighted average score will not affect the driver’s score because the passenger itself having a low rating” (SM2) | App weighs down passengers who have a lower rating | Automated decision-making |
| “What we call our hailing algorithms are based on the rating system. So, we prioritise the best rating drivers most of the time and try to encourage drivers to get a better rating. So, we give priority to the driver who has a rating above 4.7” (SM3) | App deprioritizes drivers with lower ratings | |
| “We give …. their reviews, we use that to monitor their behaviour, as behaviour of the driver if there’s any complaints, manage the complaints, make sure that we give the right feedback to the customer as well” (SM3) | Reviewing behavior of the driver and the customer | Automated evaluations |
| “So based on that loyalty … they will be getting different discounts and different partnerships” (SM3) | Benefits offered depending on ranking through app | Nudging |
| “We have certain… we got this, that part, let’s say the reward and punishing system. We did it through technology. We completely relied on the complaint mechanism” (SM2) | Rewards and sanctions through the app | |
| “We have categories for drivers now, like let’s say gold drivers, platinum drivers. That is there for some time so that also encouraged them. Plus, drivers who are doing good, they earn more money. That also makes quite a lot of encouragement for them” (SM2) | Rankings and monetary rewards for better performance | |
| “So, in our driver selection hailing algorithm, we select certain criteria, so that driver rating is one of that. So, we have given a good rating so that to make sure we pick the right drivers to the right customers and give the best experience to the customer” (SM3) | Rewarding drivers with passengers who have higher ratings | |
| “So, what we are doing, and then that’s the incentive you saw, is to make sure that they do these trips. Because if they don’t do it, we have an unsatisfied customer” (SM1) | Bonus for exceeding job quota: Incentives offered through app |
| Representative data | First-order concepts | Aggregate dimensions |
|---|---|---|
| “But obviously, I mean, we know how many hours they put, right? If we know that he’s putting twelve hours or eight hours… if it was eight hours, we know that he’s working full-time. But if he ends up putting four hours, for someone who’s, then we know there’s something going on, right? So, we just need to know how to manage that situation” (SM1) | Using information about working hours to penalize them for using other ride-sharing apps | Data and surveillance |
| “We monitor his behaviours, performance through our application, so we have built some algorithm to make sure that we capture that information. So, we periodically review those” (SM3) | Monitor and aggregate data on driver behavior for review and response | |
| “So, we’ve been monitoring the rating continuously” (SM3) | Continuous monitoring of rating | |
| “So, we have different loyalty years from a newbie to platinum, so that depending on where they land in the listing, depending on how long they have service, what is their rating, how many rides, jobs they have done…We will decide their rating, which is every month, it’ll get updated” (SM3) | Gather various information on the app to determine rating | |
| “We do a lot of changes to our technology to give a better user experience” (SM3) | Adapt technology to respond to driver | Responsiveness |
| “We give periodical feedback to the driver also” (SM3) | Feedback given to driver | |
| “We are running some machine learning and different algorithms now for handle… we have a kind of a weighted average. If let’s say the driver also can rate the customer, we look at the person. To create a passenger rating too for a driver, the weighted average score will not affect the driver’s score because the passenger itself having a low rating” (SM2) | App weighs down passengers who have a lower rating | Automated decision-making |
| “What we call our hailing algorithms are based on the rating system. So, we prioritise the best rating drivers most of the time and try to encourage drivers to get a better rating. So, we give priority to the driver who has a rating above 4.7” (SM3) | App deprioritizes drivers with lower ratings | |
| “We give …. their reviews, we use that to monitor their behaviour, as behaviour of the driver if there’s any complaints, manage the complaints, make sure that we give the right feedback to the customer as well” (SM3) | Reviewing behavior of the driver and the customer | Automated evaluations |
| “So based on that loyalty … they will be getting different discounts and different partnerships” (SM3) | Benefits offered depending on ranking through app | Nudging |
| “We have certain… we got this, that part, let’s say the reward and punishing system. We did it through technology. We completely relied on the complaint mechanism” (SM2) | Rewards and sanctions through the app | |
| “We have categories for drivers now, like let’s say gold drivers, platinum drivers. That is there for some time so that also encouraged them. Plus, drivers who are doing good, they earn more money. That also makes quite a lot of encouragement for them” (SM2) | Rankings and monetary rewards for better performance | |
| “So, in our driver selection hailing algorithm, we select certain criteria, so that driver rating is one of that. So, we have given a good rating so that to make sure we pick the right drivers to the right customers and give the best experience to the customer” (SM3) | Rewarding drivers with passengers who have higher ratings | |
| “So, what we are doing, and then that’s the incentive you saw, is to make sure that they do these trips. Because if they don’t do it, we have an unsatisfied customer” (SM1) | Bonus for exceeding job quota: Incentives offered through app |
Sharing content requires targeting cookies to be enabled. Please update your cookie preferences to use this feature.