[Record of Survival of New Employment Forms of Workers ⑥] When will the takeaway riders who are "driven" by the algorithm run their orders calmly?

  Our reporter Wang Yu, Li Runzhao and Liu Xu

  "Workers' Daily" (version 06 on November 29, 2021)

Near noon on November 24, on a street in Dongcheng District, Beijing, a deliveryman was riding a crosswalk while operating his mobile phone.

Photo by our reporter Wang Weiwei

  Obviously at the delivery location, why can't the rider press the "delivery" button?

Why is the optimal distribution plan of system planning "vulnerable" in the face of heavy snow?

In the past few days, reporters from "Workers Daily" have followed many takeaway riders to take orders and deliver meals in Beijing, Fuzhou, Shenyang and other places, and recorded their order status under the "drive" of the system and algorithm.

  Beijing takeaway rider-afraid of overtime, afraid of bad reviews, retrograde, running red light

  "How long will it take for the 35th?" "Is the 32nd ready?" At 11 o'clock on November 12, outside the Beijing Financial Street shopping mall building, more than 10 takeaway riders were waiting for a meal at a meal outlet.

Rider Liu Cong saw that the order was about to be overdue, so he didn't wait at all, and prepared to deliver the 7 meals in his hand.

  This time, he received 9 orders in total.

These 9 orders are basically within 3 kilometers, and the delivery time ranges from 30 minutes to 50 minutes.

  The time seems to be ample, but due to the concentration of orders during the noon rush hour, the time left for the rider is very limited.

"I have delivered up to 17 orders in one trip. The last 10 minutes I basically rely on sprinting, often retrograde and running red lights, or I will definitely overtime." Liu Cong said that once the timeout, the riders will face fines ranging from 3 yuan to 7 yuan per order. , "Basically equal to running in vain."

  At the same time, Li Lining, the takeaway rider in charge of the Asian Games Village area, was also busy.

Different from Liu Cong's ability to collect meals in a shopping mall, Li Lining's orders came from scattered merchants, which gave him quite a headache.

  "Different stores are far away and the speed of meal delivery varies, but the system will not consider these." Li Lining said that riders cannot refuse orders, and can transfer orders to others when they are too busy. There are 3 opportunities a day.

  “You can also choose to report, indicating that the store’s food is slow. If it is overtime, it is not the rider’s responsibility." Li Lining said, only when the rider arrives at the store 5 minutes after the order counts down, can he apply for the report. At the same time, the second time The interval between the reporting and the first time cannot be less than 10 minutes.

  The reporter saw in the backstage of the mobile phones of many riders that the distance from the rider to the merchant or from the merchant to the customer was mostly measured in a straight line.

“Obviously, you need to go to the crossroads to be delivered to the opposite side, but the system only gives you 5 minutes of food delivery time. It will not consider that you cannot cross the road directly.” Takeaway rider Wang Hui said helplessly, “During the epidemic prevention and control period, some communities did not Let the electric cars enter, they have to walk to deliver meals, so that time is even more short."

  Although the riders complained a lot about the system and algorithms, they had to compromise.

Because they are most worried about bad reviews.

  Under the logic of the big data algorithm, the delivery time and praise rate affect the "portrait copy" of each rider.

Whether it is considered a high-quality rider is all determined by the data.

Once it is judged as a "poor student" by the algorithm, the order volume and order quality will be affected.

  "The more orders are completed, the more orders will be received, which is the same as snowballing. On the contrary, if the degree of completion is poor, fewer and fewer orders will be sent, and they are all far orders, which is not thankful." Wang Hui told reporters. , These are directly linked to income.

Facing the system, "adaptation" is the only way out.

  Fuzhou takeaway rider-system positioning failure, slow meal delivery caused by the merchant overtime, but it is difficult for the rider to appeal

  Turn on the flight mode, turn off the flight mode, refresh the page, turn it on and turn it off again... At 11:50 on November 21, Fuzhou takeaway rider Liu Anmin hurriedly walked on the sidewalk of the community while carrying takeaway, while repeatedly refreshing the takeaway system page.

  Five minutes later, he was still mistakenly "frozen" by the system in another community 800 meters away from the delivery place.

At this time, a voice reminder that "you have an order is about to expire" came from the phone, and he was very anxious.

  Another 2 minutes passed, and I don’t know how many refreshes, Liu Anmin’s positioning point in the system finally "drifted" to the delivery location.

This order, he overdue again, overdue 3 minutes.

This is Liu Anmin's 17th order this month because the system positioning delay timed out.

  In Liu Anmin's mobile phone interface, the reporter saw that the end of the itinerary and the food delivery location, which should have overlapped, were separated by two blocks of buildings.

Liu Anmin sent the screenshot to the customer service.

Without waiting for the customer service to reply, he started the delivery of the next order.

  The system shows that the delivery distance of new orders is 7 kilometers away, which is 4 kilometers farther than the average distance of orders delivered by him in normal days.

"Look, this is an order, this is an overtime'ticket'." Liu Anmin said helplessly.

  Liu Anmin knows that there is a high probability that the customer service will not be able to wait for a reply that day, and even if there is, it will not be of any help to the next meal delivery.

  "Open the customer service center, and the answer will always be AI customer service; you have to wait for manual customer service, and the result will not be answered until the second day at the earliest. But the ticket will not come until 24 hours later." He said, even if there is a result, the order is overtime. It still shows timeout, and it cannot be modified or prosecuted.

  In Liu Anmin's view, the system doesn't care about the reason behind the timeout. All the rider can do is to pray that the positioning will no longer fail.

  In the background, Liu Anmin can't see the rider's rating, nor can he see the rider's level.

How much impact will the timeout have?

He didn't know, "I only know that if there are too many overtime orders, the more likely it is to be assigned to long-distance and long-term orders."

  Compared with the inaccurate positioning, Liu Anmin is more worried that the merchant’s "card meal" will cause the delivery timeout to be complained by bad reviews. "A bad review will be fined 300 yuan."

  "In order to avoid complaints from customers about the slow meal delivery, merchants sometimes click on'Already Meal' when there is no meal prepared. In this case, although the interface shows that the item is waiting to be picked up, it is actually not available. This is reasonable and reasonable. Can’t tell.” Liu Anmin told reporters that merchants can transfer the overtime responsibility to the rider, but the system cannot identify whether the merchant has a meal, and it is difficult for the rider to appeal.

  Shenyang takeaway rider—taken shortcuts by the system, overturning walls, drilling fences, and riding electric bikes on the flyover

  On November 8th, there was a heavy snowstorm in Shenyang, and the takeaway site "exploded".

  At 14:00, Zhang Jian still has 8 orders to be delivered.

The lunch box was full, and he hung the takeaway on the handlebar and clamped it between his calves, while riding the bicycle, he looked down to see if the lunch box was damaged.

Even so, reporters riding electric bikes still can't catch up with him.

  At 14:20, a familiar single sound came from the headphones.

At this time, he had 12 orders.

From the time the customer places the order, the system will dispatch the order according to the rider's location and direction, and plan the optimal delivery plan.

But what the system can’t predict is that a single snow can easily smash these-because the road is slippery, riders have to slow down, and because the platform "explodes", some orders are received late. The delivery time for the rider is less than usual.

Zhang Jian almost timed out every single order, which made him a little broken.

  After a while, Li Yong, the head of the distribution station, called to add two orders.

Zhang Jian did not dare to complain, all 23 riders in the station were on duty, and the station master also came out to deliver the order.

  "The reason for desperately sending orders is sometimes not all to make money. The system will change the method to force you to send orders." Zhang Jian said that there are more overtime orders, more complaints and bad reviews, and the site rating will drop. This means that the income of all personnel on the site will be affected.

  To fight for time, Zhang Jian modified an electric car with a top speed of 70 kilometers per hour.

When I am in a hurry to deliver meals, "the speed is so fast that my buttocks can bounce".

However, there was no heavy snow in the calf that day, which made him unable to fly even if he wanted to fly.

  "The system seems to have navigation, so we don’t need to think about the route. But if we listen to it all, we can easily be led astray and have to'take shortcuts', go retrograde, overturn walls, drill fences, and even let me ride an electric bike. Cross the street over the bridge." Every time he encounters such an order, Zhang Jian can only complain secretly, and follow suit, "the detour will definitely time out."

The system knew where the trail was opened and where the fence was demolished. The system knew earlier than some riders, and the delivery time became shorter.

  In order to fight for time, Li Yong often went to merchants with many orders to "recall orders."

Sometimes, the waiter and the chef yell a few more times to make the meal faster, but some dishes take a long time and still occupy the rider's delivery time.

"Stew has the same meal time as salad. The algorithm definitely doesn't know how to cook." Li Yong said.

  At 18:00 that day, Zhang Jian received customer complaints.

The first one is 1 kilometer and 30 minutes remaining; the second one is 2.5 kilometers and 20 minutes remaining.

To avoid overtime, he first sent the second order.

The first order customer, looking at the location, passed by his home but did not deliver food, and cancelled the order.

Zhang Jian was aggrieved.

Fortunately, the platform considered bad weather and exempted the penalty for overtime orders that day.

  At 20 o'clock, Zhang Jian's phone was called again with the lost reporter, but no one answered him.

At 1 am the next day, the shared location showed that he was parked near a restaurant.

At this time, the reporter saw Li Yong in the work group informing everyone: "No leave is allowed tomorrow. If you don't come, you will be punished as absenteeism." After that, everyone replied in unison: "Received."

  (Some interviewees have pseudonyms)