Which cases still hold up?

Audited public figures, independent customer-satisfaction rankings, and a leadership change in 2024 that deepened decentralisation. Two caveats belong in every mention.

A bank has beaten its peers for fifty years without a single budget, and when it had to save money it cut head office, not the branches.

Audited group figures, long stability, and a genuine answer to the ownership question. But it is a governance case, not a Beta structure case, and should be labelled as such.

Mondragon shows that seventy thousand people can own and govern their company democratically for seventy years - it says nothing about how they organise the work.

Strong, dated growth figures and a consistent public account of the model, but the evidence is only two to three years old and comes mostly from the company and Corporate Rebels.

Var Group abolished its org chart at the height of its growth and kept growing - the interesting question is what it looks like after its first bad year.

Real, identifiable and still active in 2026 - but small, young, layered on a conventional agile framework, without a single published organisational result, and documented only by the award body and the consultant. The 2026 change of managing director is a question mark, not a finding.

Bosch MPS shows that RenDanHeYi can be tried inside a German conglomerate - and that two years after an award for it, there is still not one published number that says what it changed.

A loss of roughly EUR 8-9m in long-term care in 2025, an intake stop in two regions since February 2026, a dynastic succession and failed subsidiaries abroad. Cite it, but never without the cracks.

Beta works at ten thousand people - as long as the payer pays for outcomes and not for hours.

Commercially and culturally strong, and the model survived the founder's death in 2022. But the decentralisation is real only at the store-operations level; the structure is a conventional retailer with an unusual leadership culture.

dm proves that treating 93,000 people as adults is good business - it does not prove that the store is the company.

Growth to 2023 is documented in Haier's audited reports and by market-share trackers; since 2024 revenue is flat and North America is Haier's stated drag in 2026. Every account of the micro-enterprise model comes from Nolan, Haier or Haier-linked authors, and GE Appliances' own 2026 communication no longer mentions it.

GE Appliances doubled under Haier - and the strongest sign that micro-enterprises were not the reason is that GE Appliances itself no longer mentions them.

Economically strong and the model survived the founder's exit in 2021. But the evidence is largely Haier's own, and there are documented control practices that sit badly with the self-organisation story.

The world's largest decentralisation was invented in China, not in Europe, and it proves that cells with their own profit and loss scale to 80,000 people - it does not prove that people like working in them.

Takeover, 100% control since 2026, investment, new hall, certification and the 87% staff vote are confirmed by Spanish press and registers. Every performance figure comes from the fund's own founders, who also run the most-read blog in the field. Cite it as a documented transformation, not as proof.

Krisos shows that a fund can buy a company, let the staff vote, remove the bosses and keep it profitable - and it shows why a movement should not be the only auditor of its own portfolio.

The acquisition and the continuing holacracy are documented; the evidence that Omron leaves it alone comes almost entirely from Luscii's own people and the Corporate Rebels network. Small, young, and now founder-less.

Luscii is the rare case of a corporation buying a boss-less company and not adding bosses - two years in, with the founder gone and every witness on the inside.

The self-management mechanics are intact and unusually concrete. But the company is private, publishes no figures, and Chris Rufer is sole owner and sole board member with no public succession.

Morning Star shows exactly how colleagues contract with each other without a boss - and it shows why a model that hangs on one owner is not yet a proof.

Citable for steward ownership and for a values-driven culture; not citable for how work is organised. The 2024 restructuring with two rounds of layoffs was ordered from the top, and the sustainability record is weaker than the reputation.

Patagonia gave the company to the planet and kept the org chart - which proves that purpose and self-organisation are two different decisions.

Citable for the Toyota Production System and for time-oriented flow; not citable for decentralised steering or for an organisation without hierarchy. The scandals show where the periphery's power ends.

Toyota proves that letting every worker stop the line beats central control - and its certification scandals prove that the same company never let anyone stop the schedule.

A profitable, listed company with micro-enterprises in its audited report - more than a self-report. But the RenDanHeYi label is months old, the effect cannot be separated from AI, price rises and headcount cuts, the family holds 57%, and no independent source has examined it. Corporate Rebels' "5,000 employees" is wrong.

VGL shows that a listed Indian retailer can put micro-enterprises into its annual report - and that even its founder says the model is halfway there.

Officially unchanged, but consistent employee reports since 2023 of a reorganisation with about 1,000 layoffs, and a Great Place to Work score down to 71%. Not verifiable from outside - the company is private.

Self-organisation outlived its founders for forty years at Gore - which is why the reports since 2023 that it is being managed back in deserve attention, not denial.

Twelve thousand jobs cut, half of management removed, a EUR 2bn savings target - and no published evidence yet of better outcomes, faster decisions or higher engagement. Cite it as an experiment in progress, never as a result.

Bayer removed half its managers and calls it shared ownership - nobody has yet shown that anything except the cost went down.

The freedom model is documented as dismantled by its own champions; the company has halved in people and revenue since 2009 and posted a loss in 2025. Cite FAVI for what happens when liberation hangs on one leader.

FAVI was free for twenty-six years and went back to classic management within a few years of its liberator leaving - liberation that hangs on one person is not a structure.

Every number is from 2021 and comes from the pilot's own leaders, Corporate Rebels or the Haier Model Institute. No evidence exists that the micro-enterprises survived the 2023-25 restructuring of Fujitsu's European business, and the leaders themselves no longer mention them. Cite it as a documented pilot, not as a case.

Fujitsu's fourteen micro-enterprises were real in 2021 - and the fact that they won an award in 2024 with 2021 numbers is the best argument for dating every case you cite.

No numbers, no structure, no follow-up: the entire Beta claim rests on a 2018 workshop and a consultancy's aggregate statement about twenty clients. Listed here so that the Corporate Rebels citation can be placed, not because there is a case to cite.

Garco10 shows how a 2018 workshop turns into a 2026 claim of radical decentralisation - the company never said so, the consultancy did, and the movement repeated it.

Every figure that makes the case is from 1982 to 2005 and comes from Semler himself. There is no operating Semco of any size today to point to. Cite the history; never say Semco does this today.

Semco proved thirty years ago that people can set their own pay and elect their bosses at 3,000 people - and then its founder sold it, which is why it is a story, not a case.

The three identity features are gone, the no-layoff record is broken, and the Beta scene has not yet acknowledged it. Citable as history (1971-2019) and as the cleanest example of what a shareholder-driven turnaround does to a culture.

Southwest put employees first for fifty years and reversed it in eighteen months - a culture without an ownership anchor lasts exactly as long as the board that believes in it.