Fusion is a coherence problem, not an org chart

When a warning is lost, the recommendation writes itself: establish a national fusion centre. It is the right instinct. But it is only the beginning of the answer, because fusion is not a place, a logo or a box added to an organisation chart. It is a property of the system – the sustained capacity for a signal held by one part to reach every part that needs it, and then the person who must act.

Services routinely stand up something called a fusion centre and remain, in practice, fragmented. The building exists; the coherence does not.

How "fusion" fails while looking fine

The failure modes are familiar to anyone who has worked inside one:

• Hoarding. The service with the best liaison reporting guards it, and the cell sees a redacted shadow of what actually exists.
• Orphaned channels. An open domestic investigation and an incoming foreign warning sit in two systems that never cross-tip.
• Passive fusion. The cell aggregates and files, but has no authority to escalate or to convert an assessment into protective action.
• Politicised tasking. Attention follows the principal's interests, not the threat, and inconvenient reporting is quietly de-prioritised.
• Depletion. An exhausted, churning analytic cadre cannot sustain the cross-reading that fusion depends on.

What real fusion requires

Coherent fusion is less about co-locating people than about guaranteeing five things hold at once:

• A shared picture – a single, current operating view that no agency can privately edit down.
• Automatic cross-tipping – a named-threat match in one stream surfaces the related case in another by default, not by favour.
• An escalation duty – a standing obligation, not a discretionary choice, to move a named-threat warning within hours.
• Insulated tasking – priorities set against the threat and protected from partisan capture.
• A sustainable cadre – enough capacity, tenure and trust that analysts can actually do the joining.

The test

There is a simple diagnostic question that cuts through the architecture debate: if a specific warning entered any node of your enterprise today, would it reach the decision-maker in time – by default, not by heroics?

Most services cannot answer that with confidence, because they have never measured it. They have measured inputs (staff, feeds, meetings) and outputs (products), but not the thing that actually failed last time: transmission. That is precisely what the Intelligence Coherence Index puts a number on – treating fusion not as a box that exists or doesn't, but as a level of coherence you can score, track and raise.