When Everyone Is Right and the Result Is Wrong

A collection of sensible departmental decisions can still produce an outcome nobody intended.

One of the more uncomfortable organisational failures is the one in which nobody has obviously failed. Engineering shows that the design met its specification, procurement that the commercial target was achieved, operations that production was protected and finance that expenditure stayed within the approved envelope. Each account is factually correct, and the organisation still ends up with a result none of those functions would have chosen had they been asked to design it as a whole.

The instinct afterwards is to search for the weak department or the person who should have intervened. Sometimes that search is justified. More often, the problem lies in the relationship between individually rational decisions rather than in one irrational decision. Each function solved the problem visible from where it stood, but the organisation does not experience those decisions separately. It experiences their combined effect.

Consider a project in which procurement is rewarded for reducing acquisition cost while engineering is responsible for reliability. Procurement selects a lower-cost alternative that meets the written specification; engineering defends a specification that reflects the minimum technical requirement. Operations then inherits an asset that is compliant and commercially attractive, but expensive to maintain under its actual duty. Nothing in this sequence requires incompetence. The weakness sits in the spaces between the questions each group was asked to answer.

Established operations show the same mechanism. Maintenance reduces backlog because backlog is measured, preventive work rises because compliance is measured, and availability is protected because production is measured. Each improvement can be real, yet the combined workload exceeds the resources available, and prioritisation migrates informally to whoever shouts loudest or whichever indicator is reviewed most visibly. The dashboard stays green while the system beneath it depends increasingly on judgement the dashboard never records.

This is why alignment is often treated too superficially. Plans, milestones and responsibilities are aligned after the important boundaries have been drawn, and a team can agree perfectly on who does what while holding different views of what matters most. Cost, reliability, schedule and safety are all desirable; their relative weight becomes visible only when one must be sacrificed for another, usually long after the meeting in which everyone agreed the objectives.

Accountability needs the same care. It is necessary, but it misleads when used as a substitute for causality. The maintenance team repairing repeated failures may be managing an old design choice; the project manager explaining a delay may be carrying an approval process designed for a different risk profile. Tracing the whole chain rarely produces a single villain or a single corrective action, which makes it harder to communicate and harder to place in an action tracker. It is nevertheless closer to how complex organisations actually create their outcomes.

The practical test is straightforward. When a result disappoints, ask each function not only whether it did its own job correctly, but what conditions its decisions created for the next one. Then ask whether its local objective would still look sensible if the person making the decision were accountable for the final system rather than the departmental measure. Strong organisations do not abolish specialised responsibility; they need it. What they avoid is confusing the success of each part with the success of the whole.