Record the evidence that would reopen a programme decision

A programme approves a phased rollout on the assumption that the service desk can support three sites at a time. Two weeks later, staffing changes reduce that capacity. The decision log still says “Approved,” and the delivery team is preparing the next wave.

The useful question is whether the evidence supporting that decision still holds. A decision record should preserve the original reasoning and identify the changes that require another assessment. Reopening an assessment does not automatically reverse approval or authorize a different commitment.

Keep the original decision reconstructable

Record the decision's scope, authority, date, options considered and evidence used. Retain conditions, dissent where appropriate and the obligations created by the chosen option. Protect sensitive material through the actual access process rather than copying it into a broadly shared programme log.

GovS 002 addresses decisions, controlled change and traceability in UK government project delivery. Its context supports retaining the basis and implications of decisions; the example record below is an original proposed method, not an official mandated template for every organization. Government Functional Standard GovS 002

The steering-paper guide concerns getting an initial decision made. This article begins after that decision, when its assumptions or conditions may change. A product requirement disagreement and a product-vision boundary have their own owners; a programme log cannot silently overrule either.

Write a trigger that somebody can observe

“Review if circumstances change” gives no one a useful instruction. In the fictional rollout, the record could say that the next wave requires an assessment if committed support capacity falls below the staffing arrangement on which the three-site plan was approved.

The trigger should point to the evidence, the person who monitors it and the next decision route. Use a definition that the team can actually apply. A proposed threshold is not meaningful if nobody knows which staff count, which hours they cover or whether they are assigned to other commitments.

This is a hypothetical record. It does not assign actual spending or release authority to a programme manager. The responsible organization must establish those permissions.

Separate the trigger from the decision

The staffing change is evidence that the original assumption may no longer hold. It does not decide whether the programme should reduce the wave, add funded support, defer a site or retain the plan with an acceptable alternative arrangement.

Assess those options and their consequences. Include existing customer commitments, operating constraints and the cost of waiting. When immediate containment is necessary, use the authority already established for it and preserve the reason. Do not rely on the existence of a decision log to create emergency powers.

If the evidence is incomplete, state that. The programme may need a bounded investigation before its decision-makers can choose. The dashboard evidence article explains why a status color cannot settle an unresolved operating question.

Supersede without erasing the earlier basis

Suppose the authorized decision reduces the next wave to two sites while a separate staffing action is completed. Create the new version with its effective scope and rationale. Identify which part of the earlier decision it supersedes and which commitments remain.

Notify the people whose work depends on the decision, using the appropriate controlled route. A revised record sitting unnoticed in a repository will not update a supplier order or a site's preparation plan. Track the required implementation and communication evidence separately from the approval itself.

Avoid rewriting version 1 to make it appear that the programme always intended two sites. That would remove the reason the team needs to understand the change. The historical record should show the information available at the time and the later evidence that changed the decision.

Review whether the trigger did useful work

After the next wave, examine the trigger as part of the programme's learning. Was it observable early enough? Did it reach someone with authority? Did the impact assessment consider the obligations created by the first decision? A trigger that fires constantly without a decision route may need redesign.

The record is useful when the team can explain why it acted under version 1, why that basis needed reassessment and what version 2 now permits. Keeping the word “Approved” visible is much less important than keeping its meaning current.

The rollout, decision identifiers and capacity assumptions are illustrative. Primary governance guidance was checked on 25 October 2026.