Figure 1 (a) captures the moment when I, as a courier, started moving towards the delivery destination after picking up the food, and the app is trying to give me a new order already—the orange notification at the top. I would open the notification, quickly scan the delivery fees, pickup and delivery locations, and the items, and hit the Accept button (or Reject if I was feeling the next suggestion would be better). All of this would be done while walking to deliver the current order. My app screen would then update with the paths for the next order to instruct my next journey (Figure 2 (b)). As this process repeated, there would always be the next order to complete. The act of turning on the New Order switch triggers ‘an automated “yes” to new orders’, which enables AI to keep feeding jobs, without a pause, to couriers on the move (i.e., couriers accept or reject the suggested jobs in motion). The practice of working with this switch turned on is colloquially referred to as Yeon-call
(meaning Continuous-orders) by the courier community. This functionality and practice is generally well-received by couriers, as they do not need to wait, even for a second, for the new order between deliveries during the peak hours. Like most other delivery platforms, Baemin pays per piece. Couriers seek to move as quickly as possible to deliver as many orders as they can in any given time. The ‘Continuous-orders’ is predicated on this much critiqued precarious piece-rate system (e.g., Cant 2020
) to govern a new urban order of ‘ceaseless flows’.
Let’s turn to how Baemin couriers’ incessant movements governed by the app materialise. To capture the spatiotemporal patterns of delivery work, I asked my courier interviewees to record each of their delivery sessions for two weeks using an activity tracking app Strava, from which I created a pack of maps and graphs that were brought to the follow-up interviews. Here I present one of the maps that visualises how the logics of the ceaseless flow at an urban scale is performed and experienced as confined movements by the courier (Figure 2).
Figure 2. Hyejin’s 11 delivery traces superimposed and animated. The distances between dots indicate speed, where the tangled dots mean slow-downs or immobile events (data gathered using Strava by Hyejin and visualisation generated using Heatflask, geographical details removed for sanitisation). [Image credit: El No]
Figure 2 captures Hyejin’s (part-time on-foot courier) 11 delivery traces in an animated form, where many paths overlap as a result of Hyejin having to walk the same paths many times. The shape of traces and the sensed quality of delivery journeys are not determined by couriers but, rather, by the platform. Another part-time on-foot courierEunyoung says: ‘The rumour says that Baemin likes to send their couriers to the places they've already been. I want to go to different places, but you know, I must go to where I’m sent to’.
It appears that smooth-flowing urban mobility envisaged by the platform requires couriers’ ceaseless movements, performed as repetitive and tethered, constrained by algorithms. Kyungsoo, describing moving with Baemin’s AI is like ‘being fish inside a fish tank’, appreciated moving along the AI-generated fences for its efficiency. Many couriers, in fact, agreed with Kyungsoo, by saying Baemin AI is ‘convenient’ as it automatically feeds jobs (Dohyeon, part-time bike courier) and is ‘trustable’ as in smart and efficient (Seongmin, part-time e-bike courier). As they did not want the ‘flow’ to be interrupted, couriers would willingly ‘give [their] bod[ies] to AI’ (Kyungsoo). On the other hand, Sangho (full-time motorbike courier) had more critical reflection, saying ‘Everything is through apps these days. It’s all contact-free, right? I sometimes have this feeling that I’m being mechanised, and I, I’m kind of a thing being moved in an app space’. Sangho’s comment demonstrates Baemin couriers doubly being stuck inside algorithmic and material urban space—with mobilities governed by the logics of Baemin's AI that render couriers as the infrastructural components mobilised to shape a cityscape of ceaseless flows.
Furthermore, the tethered movements of couriers are what enable sedentary eaters to afford immobility. Customers would acquire extended mobility through the couriers transporting food for them, keeping them safe from infectious diseases or inclement weather, while also allowing them to avoid having their daily routines interrupted. There will always be one courier who will deliver that cake and iced Americano, riding through rain or snow for an extra weather bonus. This aptly demonstrates the relationality of mobilities, where one type of mobility is interlinked with other types of mobility or someone else’s immobility (Hannam, Sheller, and Urry 2006
; Sheller 2018
). This is well documented within the critical logistics literature. Nóvoa (2014
), for example, finds that European free market of mobility requires confined movements of lorry drivers tied to a harsh geography of fixed routes and circuits. Such efforts are often invisible to customers who enjoy food, somewhat magically, arriving at their doorsteps with a few clicks on a mobile screen—‘they only need to open the door [as food will be there]’, as Junho (part-time e-bike courier) says.
As I close this post, I want to reiterate my key observations while reflecting on the broader implications of the platform-governed mobilities for urban governance and lives. Many urbanites take the convenience of food delivery apps for granted. However, we must recognise that such a smart lifestyle is sustained by couriers who perform programmed mobility, moving at the pace of the platform while staying within the algorithmically drawn fences, outside of which money does not exist. Furthermore, the platform-regulated fences block out ‘different rhythms, speeds and affects’ with which things move and vibrate (Merriman 2014
, 178). To move away from tethered mobility and towards heterogeneous and invigorating movements, rhythms, and sensations, we would need to pause the ceaselessly running wheels and, following Mimi Sheller (2018
), ask: what kind of and whose (im)mobility have we given up by pursuing a certain type of (im)mobility?
With COVID-19 and the subsequent economic downturn, many people have been pushed out into the streets, a space devoid of basic social protection. The most vulnerable are those who have no other choices but must remain imprisoned by platform-regulated mobility and immobility (cf. Massey 1994
, 149) in the longer term; whose bodies are also more malleable to technological and infrastructural changes initiated by the platform. Digital platforms generate uneven consequences in various domains, beyond the much-discussed precaritization of work, which require continuous scrutiny. I hope this post adds to that endeavour by revealing less examined, emerging type of power that platforms wield over urban spaces, mobilities, and populations.
 For context, unlike most western countries, where migrants constitute the majority of the platform workforce, in Korea, as of 2023, only those who are Korean national or have a settled status can work with delivery platforms. Baemin and other similar platforms are actively crowdsourcing the required labour from ordinary urban residents, including their customers—Baemin places freelancing courier recruitment advertisements on their customer app.
El No is a sociologist at the University of Cambridge and is a member of the Planetary Praxis research group. Her doctoral project explores how a food delivery platform orchestrates urban mobilities, bringing together spaces, infrastructures, and bodies to activate a particular kind of smartness in Seoul. Prior to her PhD, El worked in management consulting in Seoul and in analytics technology in London.