The Fix

Chapter 19: From Pretence to Autonomy

The transition is primarily political rather than structural. This chapter describes what it does to the three populations it threatens most: middle management whose coordination roles disappear, finance teams who lose narrative authority, and product managers whose competence was shaped by the dysfunction the model resolves.

The payments unit was among the first to operate autonomously. Within four months, a pattern emerged: its contract required consuming units to submit fully validated payment requests, which was structurally correct, but the consuming units faced a competing incentive to minimise their own validation costs and let the payments unit reject what it could not process.

A contract arms race followed: the payments unit tightened its schemas, the consuming units added retry logic, and payment failures increased while each unit's metrics looked healthy in isolation and the cross-unit customer journey degraded.

The contract fix was structural: the three units redesigned the contract together, splitting validation responsibilities so that each unit validated what it owned. A shared validation event allowed consumers to pre-check payment-specific fields. The arms race stopped, and the customer experience recovered within a quarter.

The autonomous-unit model had produced both the failure and the fix, but only because the people involved treated the contract as a hypothesis to be tested rather than a boundary to be defended. Scaling the model requires a transitional fund, a CTO-CEO decision pair, and a pull-based relationship between central services and the units they serve. The harder part is the human reality of that scaling, because the transition is primarily political rather than structural.

Moving from a software-dependent corporate to an organisation built around autonomous business units means dismantling structures people depend on, redistributing authority people have earned, and eliminating roles people occupy. It threatens livelihoods and identities, and most attempts fail not because the model is wrong but because the organisation underestimates the political cost of making it real.

The transition begins with a commitment to explicitness. Before any structural change, the organisation must describe itself in precise, testable terms rather than strategy language or aspirational diagrams: the core processes through which it creates value, the customer journeys that correspond to them, where the boundaries lie, which teams touch which processes, and what data flows between them. The description need not be structurally perfect, but it must be operationally honest.

In the first quarter, the work is primarily analytical. A small team, ideally led by someone who understands both the business and the systems and who has the political standing to deliver uncomfortable truths, maps the core value-creating processes. Not all of them. The three to five processes that represent the majority of customer value and organisational cost.

For a bank, this might be account opening, lending origination, payments, and customer servicing. For an insurer, it might be quote-to-bind, underwriting, claims management, and renewals. For a retailer, it might be product catalogue management, order fulfilment, returns, and customer support.

...

See also: Full contents · Preview chapters · Illusions in the Boardroom