Resilience is a cross-functional rehearsal

Technology dependencies, business priorities and communication responsibilities need to meet in a practical rehearsal.

The aim is to expose uncertainty before disruption does. Who decides which service returns first? Who understands the operational consequence of a workaround? Answering those questions turns resilience from a technical document into an enterprise capability.

Recovery plans remain untested for reasons that seem sensible at the time. Rehearsals disrupt normal work, senior leaders have crowded diaries and the risk being prepared for feels remote compared with this quarter's targets. When exercises do happen, they are often confined to technical failover tests that prove a system can be restored without asking whether the business knows what to do while it waits. Plans written mainly to satisfy an audit requirement compound the problem. They describe an orderly sequence of steps that bears little resemblance to the confusion of a real disruption.

The rehearsals I design bring business and technology leaders into the same room around a realistic scenario. The scenario unfolds in stages, with new information arriving as it would during an actual incident, and a facilitator keeps a log of every decision and who made it. I deliberately remove an assumed resource partway through, such as a key individual or a communication channel, to see how the group adapts. Each exercise ends with a review that produces specific actions, named owners and dates, which are tracked like any other operational commitment.

Imagine a scenario in which a core customer platform becomes unavailable during the busiest trading period of the year. The technology team knows how to begin restoration. The harder questions belong elsewhere. Which customers should be served first when capacity returns? Should orders be taken manually, accepting weaker controls for a period, and who has the authority to approve that? What should customers be told, by whom and at what point? Those decisions involve operations, legal, communications and finance, and they are far easier to make well if the people involved have practised them together.

The usual objection is cost: rehearsals take senior time and can look like overhead in years when nothing goes wrong. Another risk is that exercises become scripted, so participants perform a familiar routine rather than confront genuine uncertainty. I answer the first by matching the depth of rehearsal to the criticality of the service, so effort is concentrated where disruption would matter most. I answer the second by varying scenarios, introducing unannounced elements and occasionally exercising at inconvenient times. A rehearsal that feels comfortable has usually stopped teaching the organisation anything useful.

Executive committee members need to take part personally rather than sending delegates, because the decisions under rehearsal are theirs to make. That is a significant request of busy people, and it only lasts if the exercises are visibly valuable. The board also has a part to play. It should treat rehearsal findings as information that strengthens the organisation, not as evidence of failure to be minimised in the minutes. When findings are welcomed, teams report them honestly. When they are penalised, exercises are quietly designed to pass, and the organisation loses its best early warning.

The most valuable rehearsals I have led were the uncomfortable ones, where an assumed recovery path turned out to depend on a single person or an undocumented integration. Finding that in a rehearsal is cheap. Finding it during an incident is not. Boards should ask how often the organisation rehearses, and what the last rehearsal changed.