What It Looks Like

Chapter 21: What This Looks Like

A year in the life of a fictional European insurer that has completed the transition, composed from real cases. What remains difficult, what has genuinely improved, and what the people who made it happen look like on the other side.

The first thing the new engineer notices is the silence. Not literal silence; people are talking, keyboards are clicking, a standup is wrapping up in the corner. The silence is structural. No one is on a bridge call. No one is in a dependency meeting. No one is waiting for an approval that has been stuck in a governance queue for three weeks. The work is happening where the work lives, owned by the people who understand it.

She will learn, over the next three months, that this is what clarity sounds like. This chapter describes a year in the life of an organisation that has made the transition. Not a utopia but a realistic picture, with remaining tensions, trade-offs, and difficulties acknowledged. The problems change in kind: the new problems are the kind that capable people can solve. What follows is drawn from real transitions, composited and fictionalised as elsewhere in the book. Not every one succeeded fully. Some stalled when political will ran out.

A European insurance company with approximately 2,000 employees, of whom around 400 work in technology. It sells motor, home, and health insurance through direct and broker channels. Its core processes are mediated entirely through software. It is, by any measure, a software-dependent corporate.

Eighteen months ago, it completed a restructuring. The old model, a technology division serving “the business” through project-based delivery with centralised architecture governance, annual budgets, and a product management function that mediated between stakeholders and engineering, was replaced with autonomous business units organised around core processes.

The transition was a sequence of specific, uncomfortable decisions rather than a programme, each producing a reaction the next decision had to absorb.

The programme management office was dissolved in the third month. The programme director, a woman who had held cross-unit dependencies in her head for six years, was told her function was being eliminated, not repositioned. Her objection was specific and legitimate: who would track the cross-unit dependencies that her office currently managed? Without her team, she said, the first time two units needed to coordinate a regulatory change across a shared customer journey, nobody would know who owned the coordination. The CEO's response was equally specific: the contracts between units would track what the programme office tracked, and the cross-unit synthesis would surface the dependencies her team currently surfaced through meetings. The programme director understood the logic. She also understood that her twelve years of institutional knowledge about which dependencies were formal and which were informal, which escalation paths worked and which were performative, would not transfer to a contract registry. She was right about that. The organisation absorbed the cost over the following two quarters, as the contract boundaries revealed dependencies her team had been managing silently.

Two layers of middle management were removed. Three senior leaders left. Forty percent of the product management function was eliminated or repositioned.

The CFO's resistance arrived not as opposition but as a procedural objection. She presented the existing budget model: departmental reporting, quarterly variance analysis, the P&L structure the board had reviewed for a decade. The CEO presented the process-level model: cost traced to behaviour, investment connected to outcome, each unit's operational economics visible without a finance team constructing the answer from proxies. The CFO's objection was that the board expected departmental reporting and would not accept a format change mid-year. The CEO said they would present both for two quarters, until the board saw the difference. The CFO agreed. Within three quarters, the board stopped asking for the departmental view. No one had made the old process indefensible; the new one had simply made it redundant.

...

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