We continue the conversation with Callum Cant about his book Riding for Deliveroo, which, as the name suggests, documents his experience riding for the UK-based food delivery startup Deliveroo, in a bid to understand the new form of “algorithmic management” that the company represents.
In the first conversation, we started by having Callum talk in detail about what it’s actually like to ride for Deliveroo. We then went on to discuss the “workers’ inquiry” methodology and class composition analysis that inspired Callum’s research.
In this episode, we move on to discussing what lasting impact gig companies have had on the urban fabric, and explore the dissonance that has developed between the tech vision of seamless, disembodied convenience and the exploitative labour process embodied in the experiences of gig workers themselves. But first, we start by discussing the idea of invisible organisation as it relates to the two strike waves that Callum documents in the book.
Listen to the first part here.
Below is a transcript of the conversation.
Charlie Jay Clemoes (CJC): In the book, there are two spikes in strike activity that you mainly focus on, which seem to take you somewhat by surprise. Or at least they come from somewhere that you didn’t expect. In relation to that, I thought it would be good to bring in this concept of invisible organisation that you discuss. Could you just introduce the term and then also explain what was going on under the surface during your early days riding for Deliveroo?
Callum Cant (CC): I really think the important thing to understand about invisible organisation, it’s not necessarily a strategy per se, right? The term is given to us by an Italian Workerist called Romano Alquati who was observing workers in the Fiat plants and observing how they were organising to improve their situation there and how they were fighting back against the dictates of capital in their workplace and he essentially noticed that there was a very strong tendency amongst workers for them to organise not via the formal trade union channels but just to organise with one another on a one-to-one basis, and when you looked from the outside, observers could say “ohh the unions are Fiat, the workers at Fiat are very weak at the moment”. But then suddenly out of nowhere these Wildcat strikes would emerge, which would be incredibly powerful, which would shut down a plant with 100,000 workers in it. And no one could really explain this process of emergence. Because it wasn’t being done by particularly visible channels. So there’s a sense if you weren’t on the shop floor, you couldn’t really see how anything was happening. And I remember reading that shortly before I started at Deliveroo and being like, “oh, that’s really interesting. I wonder how that works in practice”. And then I go and start the job and, suddenly, you realise actually what appears on the outside like a workplace which is disorganised, which is spread out across the city, which doesn’t really have any common points of convergence. Actually there were a number of points of convergence, both: (1) Communicative, there were these mass self-communication networks which operated to allow people to talk to one another in really powerful and effective ways and were basically the networks which would call strike action (so you’re talking about Facebook groups and WhatsApp groups basically), and these were set up over periods of time through in-person connections and gradually grew and grew and grew, and provided a really febrile environment for talking about questions like “Why our wage is so low? How should we organise? How should we respond?” They have organic functions as like “Does anyone wanna play five-a-side football with my team this afternoon? But then they were also like, “This is how we’re gonna organise”. So there were those communicative points of convergence and also (2) Physical points of convergence like the zone centre, and restaurant kitchens, some restaurants will call you early for every delivery because they want to try and get the food out as soon as it is ready. They didn’t have any waiting around time. Now that for you was really negative because you’re paid per delivery. If you spend 15 minutes waiting outside their kitchen, you’re not going to be getting paid for that time. You’re gonna be getting slowed down your hourly rate is going to be negatively affected, but also the same restaurants called all of their drivers early. So you would end up with five or six of you still outside the kitchen. Because they’ve just been calling everyone really, really early and you’re spending time together. And there’s also the whole system Deliveroo works on basically having a big surplus of labour available to work around a number of surges, basically meal time surges. And that means that at 6:00 on a Friday evening, you probably don’t need that many people, but at 7:00 on a Friday evening, you need loads of people, right? Like you suddenly need way more delivery workers. So what happens is they basically built into the app this idea of the zone centre, this spare labour accumulation point where it’s like, “go here when you don’t have any deliveries to do”. And you know how restaurants are laid out in the city. They’ll often be clustered in a certain location. So you’d all wait in the zone centre just next to all these clusters of restaurants. And then when demand spikes suddenly, you’d all be in the right place to do the orders. There wouldn’t be any delay. Now that means they’ve created an informal meeting place with no surveillance over it. The app can’t do the human manager thing of “Stop talking about unions. Go back to work!” Or whatever, so they’ve just created the perfect environment, and especially when you’re there in that zone, you’re not doing a delivery, so you’re not getting paid, so the worse your wages. The fewer deliveries you’re doing per hour, the more annoyed you are, and the more time you spend in the collective meeting place where you discuss with other workers how annoyed you are, right? So through this, they created the structure of invisible organisation that if you come into it like I did, expecting disorganised workers, that wasn’t what you found. What you found was increasingly workers who had self-organised these networks and were being given more opportunities all the time to talk to one another. And that’s where the strikes emerged. I had some experience working alongside trade unions, UCU had been on strike while I was a student. I was like, “I know what a picket line is. I know how unions work, so I’ve I’ve got some vague understanding of this, we wanna try and form a union and fine, let’s start organising that”. We had a meeting, we had about 20 cyclists come together. There are probably about 400 people working in the city, so not bad, but not hugely significant. And that was the start of something we’re like, “Alright, over the next six months we’re going to build and build and build. We’re going to gain more support in the workplace. We’re going to gradually escalate”. In fact, that’s not what happened at all, because a group of Brazilian migrant workers heard that we had in fact set up a Union and they were like, “OK, alright, So what do unions do? They go on strike. We’re going on strike.” I think it was actually two weeks notice they gave us. They Informed us through the WhatsApp chats. Are we going on strike and we’re like, “I guess the union’s cool with that. Yeah, fine. Why not?” And it emerged from there. They had a chat called the Brasileroos. It’s their chat that sparked the whole thing off and that made it explode and the Union supported it, facilitated it, and carried on a series of protests after the strike, continuing the drive for the demands in the campaign and we had some moderate success.
CJC: You mentioned the zone centres which Deliveroo created at Jubilee Square in Brighton, and then there’s Ubereats which has this different but pretty similar dynamic: where by doing away with supervisors they save lots of money but then leaving that managerial function to an algorithm creates this new problem. There is this irony, isn’t there? That the strike wave occurs and it’s to do with a moment when they’re trying to hire more people. So more people come on board and that in itself creates its own contradiction. I got that feeling reading the book, you know that… at the same time that there’s this sense of despair there’s also this sudden moment of hope that the apps are seemingly unable to avoid shooting themselves in the foot.
CC: Yeah. Well, I mean, this is the thing with technological development. It’s never purely negative for workers. It’s never just taking our power away. There is no way of just changing technology purely to empower people because it will create opportunities. It may not be easy, and it may not create as many advantages as it does disadvantages. It may shift the balance of forces against our favour, but all the same, no process of development is completely negative. It will offer chances. One of the major advantages of workers’ inquiry as a method is that when you’ve got these new developments, it can be so simple to assume from the outside what the implications of the developments are. Until you actually see from the inside, then suddenly you can get a very different picture. So for instance with Deliveroo, you’ve got rid of the employment relation: we don’t have holiday pay, we don’t have sick pay, we don’t have any of that. The regulation of the relationship has totally changed. But then at the same time, that means that trade union laws have gone out the window. So like the one fundamental bit of law that prevents workers from exerting power as and when they want through their own self-organised democratic means, that’s gone. So you can organise wildcat strikes. You’re not breaking the law. The unions can support wildcat strikes. And you can call a strike with no notice. You don’t have to inform the company ahead of time. It can happen and then destroy the ability to deliver food in the evening. These developments are always double-sided, I guess that’s the thing to really stress. It’s never just positive or never just negative. And I think they really understood this because the zone centres now don’t have them anymore. They’re not a part of the app. It doesn’t say “go and wait in this location”. Because in many ways it created such a huge problem for them. Now people still wait informally in certain locations. So, for instance, the requirement to be around the restaurants during the dinner rush is no different. People will hover around the place. But in general, the creation of a specific location has changed, but it’s really hard for them because their vision for the future involves centralising stuff. You know, you have these dark kitchens on the industrial margins of cities where, again, delivery workers cluster because you might have 10 or so restaurant kitchens all operating at the same industrial park and then people will end up there. And that’s increasingly where they will talk and meet each other and put each other on WhatsApp. So even if you get rid of the zone centre, you can’t get rid of the fact that you’re moving a load of your food production into industrial spaces and in these industrial spaces, like the industrial place just behind my flat, Getir have a thing there, right? So, they have a little warehouse and a grocery delivery thing there. Then you also have all the Deliveroo riders going to the dark kitchens and you’ve got your Wagamama’s thing and you’ve got everything else, and together you see this incredible workforce that, you know, we think about workplaces in London, it being quite rare to have a semi-industrial workplace with large numbers of people apart from in like Park Royal or East London. But even in leafy South East [London], you’ve got 30-40 workers regularly assembling in large groups there. So this dream of completely siloed individual consumers and workers who will just interface seamlessly across a network, it doesn’t happen. There are clumps. There are points of convergence. There are ways in which people end up coming together because man is a very social animal. That’s what Marx says in Capital, that we do just cooperate with one another and we do end up sharing thick social ties that aren’t just dissolvable into the network of platform capitalism. There are ways that we come together and there are always alliances potentially along this stuff as well. So there’s a really interesting example that I talk about in the book, The Fast Food Shutdown, which was the first ever national strike of food delivery workers. It was done on the same day, deliberately, as the Bakers Food and Allied Workers Union had a strike of McDonald’s workers. Now Uber Eats workers do the vast majority of their deliveries from McDonald’s. So you literally have this industrial alliance between the people working in McDonald’s making the food and the people delivering the food from McDonald’s. At the same time, they’re coordinating strike action to maximise its leverage so that both of them can rebound off one another and that I think is really, really exciting and a potentially very good example of how the attempts to disaggregate people and to split us up don’t really work. And in the final instance, there will always be points of convergence around which people assemble. Now those might not be the same. Anyone looking at the nature of the urban workplace over the last five years would never say that there will always be the same huge changes in patterns of usage and stuff occurring all the time, but it does still reoccur. No matter how much you change it, fundamentally, if you’ve got a workplace where people have to cooperate and work for one another, they’re going to end up sharing spaces and also online networks where they can keep communicating even when they’re not in physical proximity.
CJC: I think it would be nice to talk about the Rebel Roo, this newspaper bulletin that was created during the strikes. And I guess how it obviously goes against the grain of this very tech-oriented platform model and I’m interested in what function it played in organising and sharing information across urban boundaries and towards more national and even international boundaries. Maybe you could also talk a little bit more about this second strike wave that occurred in 2018, also in the context of the big elephant in the room throughout the text in that it got published maybe, what, a few months before the pandemic?
CC: Ohh it’s even more so, it got published whilst I was working with Momentum on the 2019 general election. So yeah, it’s like just before the two most depressing events…
CJC: Yeah, totally. You have this little bit in it, where you’re talking about the 2017 election and the excitement of that. And it was quite sad I suppose. I mean, there’s no shortage of reminders of that euphoric moment and the other depressing side of the coin in 2019. But sorry, this is a really convoluted question, maybe just talk about the way that it went from this more direct experience of yours to something that looked like it could coordinate across national and international boundaries and the difficulties in actually making that happen.
CC: So I mean, one of the things to think about here is the fact that these platforms, as they’ve spread, have basically, because of huge amounts of venture capital funding, planted the same model over and over and over again. It’s like a cookie cutter. Every single country had the same model with some small variations on like, you know exactly how pay and hours are coordinated. But, by and large, the labour process is the same everywhere and it has spread really rapidly. So if you think about it in terms of. There’s a good analogy from biology I remember from I can’t remember, my GCSEs or whatever, you know the idea of a monoculture, right? If you grow loads of the same plant that are all genetically very similar to one another, then one disease can come along and it wipes it all out. Well, essentially what they’ve been doing is the industrial equivalent of a monoculture. They’ve been putting the same labour-process out in all these different places with extremely minor adaptations because they have this move fast and break things approach where they aren’t going to try and fit in with local regulation, they’re just going to try and do what they want to do. So, when we were looking at the situation in the UK and then you talked to a worker in France, it would be fundamentally the same. You talk to work in Germany, it’s fundamentally the same. Whereas if you have an international supply chain, all the points along that chain, people may not understand how the labour fits together or what interests they have in common or there are all sorts of these potential frictions. But for us, there weren’t really: you work for Deliveroo. You work for Deliveroo, you work for Deliveroo, you work for Uber Eats. We all work for these platforms. It’s all the same platform. We all have broadly the same antagonisms across a whole range of different labour markets. And now you see the same thing in South America. We all work for Rappi or we all work for Glover or whatever. So by rapidly reproducing the same model over and over and over again, they put loads of people in the same position. And what’s really interesting for me is you look back at the history of capitalism, often, capitalism has tended towards the creation of differentiations within a workforce, not for any particular practical reason in terms of production, but just to stop that workforce realising they have something in common. So, Kathleen Stone wrote in the 1970s, a brilliant article in Medical America about the development of the American steel industry and basically how they were two job roles, there was semi-skilled and there was skilled. Everyone in all the steel plants fitted into these two job roles. But they realised over time that, “Oh my God, this makes it really easy for them to unionise. So, in fact, what we need to do these early unions (I think the early American Steel Union was called the Brothers of Vulcan, which may, I mean, obviously gender aside may stand up as one of the best union names in the history of the movement) they could basically organise everyone in these big camps. To combat this, the employers intentionally introduced job differentiations. If you now look at a car plant in the UK, you can have up to 1000 different pay structures and this is a plant of four to five thousand 000 workers, right? So in your plant, you may share your pay structure with four other people on average and who you’re working with on your team may have entirely different wage differentials: “I’ve got a bonus of value this… I have a bonus if I do that”. So people get split up into these groups and deliberately by capital, but they failed to do that here and that produced essentially a giant recomposition of lots of different workers around the place.
CJC: I was just reminded, it’s a little bit like this concept of the elimination of space by time that David Harvey talks about, that there’s this time-space compression going on where everything’s happening at the same sort of pace and there really isn’t much differentiation between different places, which I’ve never really heard it talked about as something that could actually create the conditions for the disintegration of the system of control entailed within the labour process. I don’t know if I got that right…?
CC: No, no, absolutely. I mean, it was really amazing in 2018 when we were doing the early versions of the Rebel Roo. So you’d be talking to someone in France about what their strike tactics were and you’d be taking the best bits. You’d be bringing it back here and then you’d be doing it in the UK. And then you talk to somebody in Germany and we were constantly in this community process where people were finding each other for the first time and you could just see month on month. They’d be like, “Oh, there was a little bit in Bologna here, there was a little bit in London, there’s a little in Brighton. Ohh, now it’s Berlin now it’s now it’s Barcelona, Madrid. Now it’s there”. There was one period I remember where it was all the Bs: Berlin, Barcelona, Bologna, Brussels. Everyone in the B cities was on strike. And this kind of development was really remarkable. Because it it had a lot of potential and I think it still does have a lot of potential. I’m not as involved in the organising side as I was, but workers from all these different contexts do have shared interests. Still, what’s interesting now is the development of regulation. So for instance, Spain now has its Riders Law, which has started to break up this common, because now you have much more varied conditions and in some circumstances like Deliveroo just left Germany. They’re just like “They’re gonna regulate us too much, we’re just gonna leave”. So that was perhaps one of the affordances of a particular moment in time. But just on the Rebel Roo project in general, I thought it was fascinating that all it did was share ideas. It was written in this way that was a bit aggy, a bit funny and written right at the forward edge of what people were willing to accept. So it would be like “Fuck the bosses!” Basically, “fuck the bosses” and then details about what other people were doing, information about what other people were doing. Once you combined all that it was enough. People were writing stuff and it flew in and we distributed it and the distribution network grew ridiculously. I mean it’s just two sides of A4 that you hand out that have like you know six or seven translated introductions of what the thing was on the right and then some blocks of text and all you had to do was put this e-mail address up and people would email asking for it. We had people distributing Glasgow and all the rest of it, even in cities where you have no contact with anyone else, like, you know, there’s someone in Canterbury who wants 300 copies and you’d go “OK. Fine, sure. Go for it.” And so it spread rapidly and it was a really, really interesting process. I think what it really demonstrated to me was this absolute thirst for information that workers had about the workplace and this is something that we talked about, with the Johnson-Forest Tendency earlier, the Johnson-Forest Tendency wrote about the way that workers in car plants used to get their mate to cover them for a bit, and then they’d go off and walk around the car plants and try and understand how it worked and like “how does my labour fit into the general social labour?” I think that drive is still really, really strong. People want to understand. And this is one of the challenges of platform management is that it black boxes or it it hides a lot of the process from workers themselves. It makes them information-poor in ways that empower management. Historically, this is what’s known as Taylorism, one of the original schools of management. Frederick Winslow Taylor’s inaugural bastard of the whole thing. He came up with a way of studying work in order to get information on the labour process that would allow managers to cut out the empty spaces that workers were preserving to prevent work intensification. He was like “I can make them work harder. I understand the process just as well as they do, so I can see all the gaps that they’ve been slowing down in and I’ll eliminate all the dead time”. And that was what the process has always been for managers: “I need to gain the understanding of work that my workers already have. I need to get that knowledge, but then with black boxing it almost in some ways reverses this, as in, the workers don’t understand their own labour process and so they have to try and steal information back from the manager. When new technology is introduced, especially stuff that hides things and that’s deliberately obscure and unexplained, workers have to engage in this struggle to bring information back, and it’s one of the elements of the Spanish Riders’ Law we were just talking about, it allows trade unions to request explanations of how algorithmic management is implemented. So this battle over information is, I think one of the really interesting elements as well here, where workers need to understand what’s going on in order to be able to fight it effectively. So the Workers’ Inquiry in that regard isn’t just about finding out what workers already know, it might also now consist of what can we find out that we can inform workers about, which can, in turn, allow them to take further steps or what can workers find out themselves that allow them to? Take further steps. A researcher would say this, but information is one of the key weapons of war in this stuff.
CJC: It really does feel like a bit of a mystery, the black box, the way that you describe it, that it demonstrates this idea of the rules run amok, that they operate without any rhyme or reason to them, which I guess to someone experiencing it must be so scintillating, just the question of “What is it all about? Why has this happened?” It must be so enticing to want to know.
CC: And people came up with a craziest theories man. People came up with the wildest conspiracy theories because, and this is where some of the really negative mental health effects of algorithmic management and the implementation of AI in the workplace more generally become really obvious, you need to understand the logic of what you’re doing. A Victorian method of punishment was the treadmill. You did pointless work endlessly, right? Not understanding the end goal or the contribution that your labour is making to society makes it fundamentally impossible to understand. And what your role in the world is. It massively reduces job quality. So, if you are fundamentally depriving workers of information about how their work works, for want of a better phrase, you are essentially contributing to a process that is guaranteed to reduce things like their self-worth, their understanding of what the importance of their labour is. When we’re looking for this information, we’re not necessarily just trying to find out the things that can help workers struggle better, we’re also trying to make sense of like “What did all these deliveries mean? How did they work? “hat was going on in my life in that period when I was doing all those things?” And for people who still work there, these are live questions, you know, “why did I get this order and not my mate?” It’s not a minor question. It’s a really fundamental one. If people want to understand and have dignity in their labour, it’s not something that can just be brushed off as like, “oh, this is intellectual property”. It’s like, “no, you’re controlling someones life. You need to explain how that’s happening”. This is just part of the class structure of production: workers have always been forced to do things that they don’t quite understand. And there’s always been attempts to obscure basic information from people. But there’s a bit in the book that goes into what the potential social use of Deliveroo might be like, what a positive vision might be, and that’s where the information stuff is so much clearer. You could understand why you’re working for social good. And I can imagine in a different society the knowledge of what you’re doing and that drive to understand could have a much more productive outcome than just feeling frustrated at an app for not telling you why you got the order.
CJC: Yeah, you talk about the social reproductive function of Deliveroo and that it does provide a pretty important service in reproducing the white-collar workforce that I suppose is maybe becoming increasingly, I don’t know, maybe not proletarianised, but certainly…
CC: Definitely some parts are. I mean you wouldn’t say it’s an even trend across the entire white-collar workforce, but definitely, some parts are becoming proletarianised.
CJC: …Just coming home totally exhausted and maybe also living in a flatshare rather than in a nuclear family, so not really being able to rely on this homemaker to cook your food and take care of you and provide a comfy house to recuperate in. Instead, you come back exhausted and have a little slice of comfort in front of your Netflix as you quite vividly portray.
CC: Yeah. So the number of people who have been like, yeah, that was me. I didn’t want to realise but that’s what I was doing. I was like, “Sorry I didn’t mean to…”.
CJC: But I was gonna say something else about this mysterious quality of the black box as well, so you already kind of were touching upon it.
CC: Well, I think there’s something really interesting about the spatial side of this whole thing, which is that what you’re really talking about with a lot of platform stuff is last-mile delivery services. And the way in which last-mile delivery services are recomposing urban environments is really, really, really fascinating. You basically have local high streets and last-mile delivery services and actually your retail park thing is being killed by e-commerce and that restructuring is profoundly interesting because in some ways people are spending more time on local high streets, potentially, but then also they’re evacuating previous shared spaces. So, for me, when I’m working at home, there is absolutely no reason apart from seeing friends for me to leave a mile radius of my flat. Because I can go to the local high street and I can get everything else delivered. There is no need. So I think that the way that last mile plays into this, like ongoing recomposition that’s introduced by e-commerce and exaggerated by the pandemic, is really interesting. For me, it’s a question of like, “How far is this a genuine siloisation? How far am I only seeing this from my working-from-home perspective? And how far is this an actual major process that entails certain kinds of urban spaces like Surrey Quays shopping market dying? I don’t know. But the Lewisham Shopping Centre in Lewisham town centre is empty now. Like, properly empty. The only things that have people going into them are the food shops.
CJC: No way. I had no idea about Lewisham Shopping Centre. It’s been a while since I’ve been there. Like five years or so.
CC: Yeah, well, I lived in the little flat opposite the clock tower during the pandemic. It’s amazing. You think of the shopping mall as one of the emblematic spaces of Blairite Britain: the idea that when we were kids, we would just go and hang out in the shopping centre. But now those things are so dead and everything’s recomposing around experiential service commodities like nice coffee places or whatever. And maybe shops that are kind of nice to shop in person, that look slightly arty. Everything’s like Lordship Lane. Everything’s like Lordship Lane or it’s an ALDI or it’s getting delivered. That’s basically my perception of London now.
CJC: Yeah, yeah, yeah. Which is a very rapid thing. It’s a very recent thing, and it’s happening right under our noses. I guess this wave of expansion of delivery services in its wake, has reconstituted the city and it has created new needs or new wants more like.
CC: So because need is social, what people need to consume always varies historically. It’s like that Marx thing about social needs now requires cigarettes and newspapers, whereas it didn’t during the mediaeval period, right? The like, these things become needs once they’ve been offered.
CJC: It reminds me of this article that Edward Ongweso Jr. wrote for Vice Magazine earlier in 2022, “How One Billionaire With a 300-Year Plan Fueled the Popping Tech Bubble,” I don’t know if you had a chance to read it. He talks towards the end about “the full picture… the desperate attempts of Uber and other app-based labor platforms like it to realize profitability [and how these have] have real consequences [and then he lists]: experiments over the years [that] have locked out tens of thousands of drivers, left tens of thousands more to fend for themselves in persistently declining working conditions, unleashed racist algorithms that discriminate against drivers and riders, and restructured life in countless cities.” It’s quite a stirring conclusion, the way that all this money that flooded into the industry has had these irreversible consequences for what we expect from an economy. And that reminds me there was another article doing the rounds at the same time, “A Nation on Hold Wants to Speak With a Manager” by Sarah Lyall for the New York Times. It was a very interesting report on the way that people have come to expect a seamless service and how that came up against the reality of pandemic restrictions, and it brought up the question of like “is this really a need, do we really need to have something in five minutes flat?”. It’s now created an impossible yardsick given what’s coming in the future.
CC: My current book is quite a lot about climate change and it’s really interesting to think about what happens when we are forced to reverse a lot of these things. What happens to the long-term reversal of just-in-time production because resilience and global shocks don’t allow it? Your back to “you can buy the car that’s in the shop and no other”. The experience of the consumer is really going to change there. I always think of the analogy to the Trabant and Soviet production and how unsatisfying the lack of washing machines was for the Soviet consumer. Well, if you think that was unsatisfying, you wait until we go from the living standards of 2019 to the living standards of 2040 and you see what happens to the Global North consumer in that process. Resilience and the creation of resilience in global supply chains basically means going back to that much more centrally coordinated push production rather than pull production. This will totally change the viability of consumer goods like consumer durables just won’t operate in the way they do currently.