e driver is supposed to make the delivery, and the driver is typically afraid of what the customer will do if the delivery isn’t on time. So when investors on Wall Street and their equivalents in Hong Kong, Singapore and other financial hubs listen to retailer and superapp executives waxing proudly about how their e-commerce delivery times are supercharging growth, they are largely or completely unaware of the true social costs of those ever-faster delivery times. Survey after survey of motorcycle delivery riders over the past few years — motorcycles are the principal mode of delivery for the e-commerce and food delivery business across much of the region — have supported what an observer can see with the naked eye just by watching driving behaviours on Asian roads: the online delivery business is dangerous as hell, both for the riders themselves and for those with whom they share the road. And customer expectations are a big part of the problem. Malaysia is a particularly dangerous place to be a delivery driver: a 2022 survey of 225 motorcycle food delivery drivers (MFDRs) by researchers in that country found that 93 of them (more than 40 per cent) had been in crashes while making deliveries. Of these, 27 had been involved in more than three accidents. Studies in China, Singapore and Thailand show similar rates of accidents. What is going on? And why are customer behaviours such a significant contributing factor? Carnage in Thailand A good place to start looking for an answer is in neighbouring Thailand. Medical researchers at Chiang Mai University in Thailand recently produced a study based on the results of a survey of more than 700 MFDRs in Chiang Mai Province who are employed primarily by Grab, Lineman and Foodpanda. (Note that Foodpanda has quit the Thai market altogether since this survey was done.) The city of Chiang Mai itself is large: it is the second-largest in Thailand after Bangkok, with a population of 1.2 million. The province in which it is the capital has a population of nearly 2 million. It is a major tourism destination and also has a large expat population. Owing to its size, demographic mix and, like just about every other major Thai city and town, the congestion on its roads, Chiang Mai is an ideal location to survey MFDRs. The Chiang Mai study examines injuries and accidents to MFDRs during Covid and how these relate to self-admitted “risky riding behaviours”. It should be noted that during Covid there was usually less traffic on the roads, so the results would be conservative compared to what the same survey might turn up based on if it were done today. Respondents were asked why they engaged in the risky behaviours in the first place. The study found that over six months during Covid, just under 20 per cent of the riders had been involved in at least one traffic accident. (The average number of accidents was 2.44 per rider.) Of these accidents, 14 per cent caused major injuries resulting in ER visits or hospitalisation. As concerning as these numbers are, they don’t fully reflect the risks taken to incur the accidents, the near misses, or the risks to other drivers. 84 per cent of the riders admitted to engaging in at least one out of 13 risky behaviours and 6 per cent engaged in more than 10 such behaviours. The behaviours included using a hand-held phone while driving, riding in the wrong direction and/or on the sidewalk, running red lights, not using indicators, unsafe lane-splitting and speeding. The results, which are pretty awful, are not at all unique to Thailand: Vietnam and China are worse; Cambodia, where anarchy on the roads goes unchecked by the authorities, is probably worse, too; and Malaysia, Singapore and Indonesia are all up there. Why drivers take risks Why do the drivers say they take engage in these unsafe behaviours? Road conditions (90 per cent of the riders), weather conditions (79 per cent), vehicle conditions (72 per cent), wayfinding (78 per cent) and customer behaviours (72 per cent). While the first four of these are not surprising, the last one is genuinely eye-catching. MFDRs are afraid of what their customers will say or do if they are late on an advertised delivery time. And with good reason: anyone who has seen unhappy customers screaming at or otherwise intimidating delivery drivers for, say, coming to the wrong building entrance, knows exactly what this means. Consider what happens when a food order is placed with Grab, the most popular superapp. The technology behind the app judges the distance between the pickup point and the destination, and has to estimate how long the order will take to prepare, and how long it will take to deliver, given the distance and prevailing traffic conditions. It uses these estimates to provide the customer with a delivery time within an approximately 10-minute window. There is a lot of uncertainty built into each of these calculations, so the app updates the delivery time window when necessary. Even so, the real-world conditions and the need on the part of the driver to complete as many deliveries as possible during his/her shift — the commissions on each delivery are extremely slender — place multiple incentives on the driver to take risks. When the driver does arrive, they confront multiple risks at the destination too: the customer may not be ready to take the delivery, or fail to communicate complicated delivery requirements, or may be in a building with more than one entrance, or a complex with more than one building, or just be impatient or abusive. It isn’t just food delivery Coupang in Korea touts delivery speed as one of its major selling points, while global retailers like Walmart and Amazon are quick to inform the investment community about their own delivery speed metrics. For example, half of Walmart’s sales in China are made online, and about one-third of the deliveries are now made within one to three hours. Walmart partners with a local Chinese company, Meituan, to make deliveries inside 30 minutes and Meituan uses motorcycles. It doesn’t take much imagination to envisage the problems a 30-minute delivery commitment on dodgy, congested roads by motorcycle in a Chinese city with its massive high-density residential buildings might cause accidents. Customers need to chill Retailers and superapps have created unrealistic expectations that commit their delivery drivers to risky behaviours and even then face the wrath of customers who don’t seem to be able to empathise with them for falling short, and who are often themselves responsible for heaping stress on the delivery person by communicating poorly, omitting key information, not being ready, or blowing their tops at the rider because they expect perfection or because they’ve had a bad day. With technology now routing deliverers to do multiple orders on a single run, every time a customer doesn’t do their bit, the rider is behind schedule and immediately in a situation where they’re more apt to take risks. Wall Street may not care, of course, but we should. Further reading: How the push for profit is causing delivery drivers to compromise safety