Whenever self-management comes up in a boardroom, someone offers the same rebuttal. It sounds fine for a design studio or a software team, they say, but real industry needs a chain of command. Machines, safety, throughput, and capital do not run themselves. It is a confident objection, and the two hardest cases in the field flatly contradict it.
Tomatoes at industrial scale
Morning Star is not a boutique. Hamel (2011) documents it as the largest tomato processor in the world, responsible for somewhere between a quarter and a third of all the tomatoes processed each year in the United States, with revenue above 700 million dollars in 2010 and around 400 full-time employees. This is heavy, seasonal, capital-intensive food manufacturing, the opposite of a nimble services firm. And it has run for more than two decades with no managers, no titles, and no promotions, with compensation decisions made by peers.
The coordinating mechanism is not charisma or good vibes. Hamel (2011) describes the Colleague Letter of Understanding, a written agreement in which each employee spells out their commitments to the specific colleagues their work affects. In a plant where one process feeds directly into the next, that document does the job a supervisor would otherwise do, aligning who owes what to whom. The founder, Chris Rufer, has claimed that volumes, revenue, and profit grew at double-digit rates over twenty years against an industry averaging around one percent a year, with growth financed almost entirely from internal sources. That figure comes from the founder inside a case study, so it should be read as a self-reported claim rather than an audited result, but the operational fact stands: a major industrial processor runs without managers.
Appliances at conglomerate scale
If a tomato plant is not industrial enough, consider a home appliance giant. Kanter and Dai (2018) document Haier's restructuring under Zhang Ruimin into market-facing microenterprises spread across five business platforms, from its core white-goods manufacturing to incubation, finance, and beyond. The crucial design choice, in their account, is that the platform owners above these units were explicitly not their supervisors. A manufacturer producing refrigerators and washing machines at global volume chose to route authority to small entrepreneurial units rather than a traditional command hierarchy, chasing what Zhang called zero distance between employee and customer.
Haier is a teaching case, not a controlled study, so the honest claim is narrow: this is a proof that a large manufacturer attempted deep self-management, not a demonstration that it beats the conventional alternative on every metric. But that narrow claim is enough to sink the services-only objection. Two of the most radical bossless structures on record belong to a food processor and an appliance maker, not to a startup.
The scale of both cases is the point. Hamel (2011) puts Morning Star at around 400 full-time employees and more than 700 million dollars of revenue, processing between a quarter and a third of the United States tomato crop. Kanter and Dai (2018) place Haier's microenterprise experiment across five business platforms inside one of the world's largest appliance makers. Neither is a boutique that could plausibly run on informal goodwill. They are large enough that coordination genuinely has to be engineered, which is exactly why the objection assumes they should not work, and exactly why it matters that they do.
Why the frontline objection gets it backward
The usual worry is that industrial work is too coupled and too risky for autonomy. The cases suggest the opposite reading. Martela and Nandram (2025), in their study of Buurtzorg, identify low interdependence between units as a condition that helps self-management, and both Morning Star and Haier engineered their structures precisely to reduce coupling: Morning Star through explicit peer contracts that make interfaces visible, Haier by turning divisions into units that face the market instead of each other. In other words, industry is not automatically too interdependent for self-management. It can be deliberately decoupled so that it is not.
That reframes the practical question. The issue is not services versus manufacturing. It is whether the work can be organized so that the people on the frontline have both the information and the authority to act, and whether the interfaces between units are made explicit rather than mediated by a boss. Hamel (2011) shows one way to do that with written agreements. Kanter and Dai (2018) show another with market-facing units. Martela and Nandram (2025) show a third with small self-contained teams. None of the three is a services firm exploiting some special softness in its work.
There is a further wrinkle that the services-only crowd tends to miss. Scale and capital intensity, far from ruling self-management out, can be part of what makes the peer mechanisms necessary and therefore workable. In a tiny shop everyone can coordinate by talking across the room, so nothing formal is needed and nothing formal gets built. In a plant the size of Morning Star, the interfaces between people are too many to hold in anyone's head, which is precisely why Hamel (2011) found the coordination written down in explicit agreements. The formality that skeptics assume requires a boss is, in these cases, what let the firms do without one.
For a manufacturing leader, the takeaway is not to abolish supervision this quarter. It is to stop treating the shop floor as inherently unfit for autonomy. The two most cited industrial cases in the literature took the opposite bet and made it structural. The real constraint is design effort, not the nature of physical production. Where the interfaces between people can be made explicit, whether through contracts, market-facing units, or self-contained teams, the supervisory layer is doing less than the objection assumes, and a food processor and an appliance giant have already proved the point at scale.
Sources
- Hamel, G. (2011). First, Let's Fire All the Managers. Documents Morning Star as the world's largest tomato processor running for two decades without managers, with the CLOU mechanism and the founder's self-reported growth claim. https://hbr.org/2011/12/first-lets-fire-all-the-managers
- Kanter, R. M. and Dai, N. H. (2018). Haier: Incubating Entrepreneurs in a Chinese Giant. Shows a global appliance manufacturer restructured into non-supervised, market-facing microenterprises. https://www.hbs.edu/faculty/Pages/item.aspx?num=54098
- Martela, F. and Nandram, S. (2025). Buurtzorg: scaling up an organization with hundreds of self-managing teams but no middle managers. Supplies the low-interdependence boundary condition used to explain why industry can be decoupled for autonomy. https://link.springer.com/article/10.1007/s41469-024-00184-y