The current fashion for delayering usually means one thing: a reorganization that removes a tier of managers, reshuffles reporting lines, and leaves the underlying machine intact. A Chinese appliance maker attempted something more radical and more durable, and Harvard has spent enough time studying it that the experiment can be discussed with a straight face.
A conglomerate rebuilt as a market
Haier, one of the world's largest home appliance manufacturers, has restructured itself under a model its chief executive Zhang Ruimin calls RenDanHeYi. Kanter and Dai (2018), in a Harvard Business School teaching case, document the shape of it. The company was organized around business platforms spanning five sectors, including its white-goods transformation, an investment and incubation arm, financial holdings, real estate, and a cultural industry unit. On top of those platforms, Haier incubated microenterprises, small market-facing units expected to earn their keep from customers rather than from an internal boss's approval.
The telling detail in Kanter and Dai (2018) is that the platform owners were explicitly not acting as supervisors of the microenterprises beneath them. That single design choice inverts the logic of the traditional conglomerate. In most large firms, a division head exists to allocate, monitor, and discipline the units below. At Haier, the platform is described as a place to incubate rather than a chair from which to command. Zhang's stated ambition, in the language of the case, was to eliminate the internal walls of the company and achieve what he called zero distance between the employee and the user, with employees and customers pulled toward the position of shareholders rather than subordinates.
Because this is a teaching case rather than a peer-reviewed study, it is worth being precise about what it can and cannot support. It documents a structure and a philosophy in detail. It does not, and this piece will not, turn that into a claim that the model provably outperforms conventional management. The value of Haier here is as a proof of possibility, evidence that a manufacturer at serious scale chose to route authority to small units facing the market, not as a controlled result.
The breadth of the five platforms in Kanter and Dai's (2018) account is itself part of the argument. This is not a single product line experimenting at the edges. The structure spanned the core appliance business, an investment and incubation arm, financial holdings, real estate, and a cultural industry unit, which means the microenterprise logic was applied across genuinely different kinds of work rather than confined to a safe corner. A company willing to run that logic across finance and manufacturing alike is making a bet about how organizations should be governed, not merely tolerating an internal skunkworks.
Same principle, opposite direction of travel
Set Haier beside the two most cited Western cases and a pattern appears. Buurtzorg, as Martela and Nandram (2025) describe, grew self-management from the ground up: nurses forming small autonomous teams, with headquarters deliberately kept thin at two directors and a ring of coaches. Morning Star, in Hamel's (2011) account, dispensed with managers entirely and let colleagues negotiate their commitments directly. Both are bottom-up stories, built by practitioners inside relatively self-contained work.
Haier is the counterexample that matters, because it is top-down. Zhang Ruimin drove the transformation from the chief executive's office, deliberately dismantling the hierarchy he presided over. That distinction should reassure and warn in equal measure. It reassures because it shows self-management is not confined to founder-led boutiques or nimble service firms; a legacy industrial giant can pursue it. It warns because a model imposed from the top depends heavily on the person at the top continuing to want it. Buurtzorg's design is embedded in how the teams are constituted. A microenterprise structure that a chief executive installed can, in principle, be reversed by the next one.
What leaders should actually take from it
The usable lesson is not the vocabulary. Copying the words microenterprise or zero distance changes nothing. The lesson is the mechanism Kanter and Dai (2018) put at the center: the unit is made to face the customer directly, and the layer above it is redefined so that it incubates rather than supervises. That is a concrete, checkable design decision. A manager can ask whether their reports serve an external market or an internal approver, and whether the level above them adds capability or merely adds oversight.
Hamel (2011) makes the parallel point from the American side, arguing that the cost of management is not only salaries but the tax that layers of approval levy on speed and initiative. Haier's answer was to make the market, not the manager, the source of discipline. Martela and Nandram (2025) show the Dutch answer was to make the team, not the supervisor, the unit of accountability. The three cases sit on three continents and use three different grammars, yet they converge on the same sentence: push the decision to whoever is closest to the value, and rebuild the surrounding structure so that it supports that person instead of second-guessing them. Delayering that stops at removing a tier of managers misses the harder half of the job, which is redefining what the remaining tiers are for.
Sources
- Kanter, R. M. and Dai, N. H. (2018). Haier: Incubating Entrepreneurs in a Chinese Giant. HBS teaching case documenting the RenDanHeYi model, five business platforms, non-supervisory platform owners, and Zhang Ruimin's zero-distance philosophy. https://www.hbs.edu/faculty/Pages/item.aspx?num=54098
- Hamel, G. (2011). First, Let's Fire All the Managers. Supplies the argument that management is a cost and the Morning Star contrast used to frame Haier as a top-down case. https://hbr.org/2011/12/first-lets-fire-all-the-managers
- Martela, F. and Nandram, S. (2025). Buurtzorg: scaling up an organization with hundreds of self-managing teams but no middle managers. Provides the bottom-up Dutch counterpart and its thin-headquarters design used for the direction-of-travel contrast. https://link.springer.com/article/10.1007/s41469-024-00184-y