When does a powerful coalition become an operating model?

When Amazon, Berkshire Hathaway, and JPMorgan Chase announced a healthcare collaboration in January 2018, the combination itself attracted attention. Each organization had resources, employees, and experience operating at scale. The announcement promised an independent company focused initially on improving healthcare for their United States employees and families. It also acknowledged that the effort was at an early planning stage. The distinction is important: an unusually powerful coalition had been formed, but the announcement was not yet a complete operating model.

The sponsors described goals that many employers would recognize: better satisfaction, simpler access, transparency, quality, and lower costs. Technology was part of the proposed approach. Those objectives were compatible at a high level, but they did not specify a single intervention or a shared purchasing decision. Reuters' contemporaneous reporting captured the market reaction and the uncertainty around what the venture would actually do. The public excitement therefore preceded much of the evidence that would be needed to evaluate a particular solution.

The venture became known as Haven and recruited surgeon and author Atul Gawande as its leader. Yet a respected clinician and prominent sponsors could not remove the work of choosing among problems. An employer might prioritize benefit navigation, drug costs, primary care, or the predictability of out-of-pocket spending. Each choice requires different contracts, partners, data, and implementation authority. These are analytical implications of the venture's stated scope, not claims about the private debates or motivations of its executives.

In January 2021, reporting on Haven's announcement said it would end independent operations at the end of February, after roughly three years. Banking Dive described the original goals and the decision to close. That establishes a bounded outcome: this joint organization did not continue in its original form. It does not establish that all its work was worthless, that the sponsors abandoned healthcare, or that a single obstacle explains the decision. Those stronger conclusions would require evidence beyond the public closure announcement.

The case is useful because a coalition can possess influence without possessing a unified workflow. Employees receive care through providers, insurance arrangements, pharmacies, and local markets. A sponsor's ability to fund a project does not necessarily give the project authority to change each of those relationships. The relevant question is where the coalition can make a concrete decision that alters an outcome. Without that specificity, a large mandate can produce many plausible projects without a clear way to choose, compare, or scale them.

This is not evidence that healthcare is uniquely immune to improvement, or that collaboration among large organizations must fail. It is a reminder that complementary reputations are not the same as complementary operating commitments. A useful partnership defines what each participant will contribute, what decision it is prepared to change, and how disagreements will be resolved. It also chooses a measure of success that belongs to the intervention, rather than borrowing confidence from the scale or accomplishments of the institutions behind it.

An ecosystem leader can begin by making the coalition's authority visible around one shared problem. Identify the affected population, the service that will change, the contract owner, the implementation team, and the result that would justify another step. A small test can reveal whether the sponsors' interests align in practice. Haven's ambition was understandable. Its teaching value lies in the distance between wanting a better system together and building an organization capable of making a specific part of that system work differently. The distinction becomes clearer when every partner names the decision it is prepared to change.

For a discussion tailored to your team, explore a workshop with Julien.