The plan uses every available hour. Then one client needs a correction and an essential colleague is absent. Keeping capacity uncommitted is a deliberate decision about the uncertainty your service must absorb. There is no universal reserve percentage that makes every team dependable.
Photo by Cabri Caldwell on Unsplash.
Use your normal capacity plan to see committed demand. Then decide which uncertainty needs protection before accepting another promise. This article addresses that commitment decision rather than rebuilding the planning worksheet.
Identify what makes demand uncertain
Review recent periods when the plan failed. Was the cause unpredictable arrival volume, variable case complexity, rework, support demand or unavailable specialist review? Some recurring work should be included as ordinary demand rather than repeatedly described as an emergency.
Separate uncertainty from missing information. If the team has not estimated a known assignment, calling the empty space a buffer hides unfinished planning. Improve the estimate and show its range.
Write a reserve decision note
Before accepting new work, record:
The service commitment the reserve protects.
Observed sources of variability and limitations in the evidence.
Role and time window that need protection.
What may use the reserve and who authorises it.
Point at which unused reserve can be released, if appropriate.
Options when uncertainty exceeds the planned protection.
Reserve must sit with a relevant capability. Unallocated administrative time will not resolve a shortage of specialist approval tomorrow. Avoid treating every spare hour as interchangeable.
Test the next commitment against a plausible disruption
In an illustrative agency, the senior reviewer has a small uncommitted window before several client releases. The owner considers selling it to a new assignment. Recent work shows that final corrections often require that same reviewer, although the amount varies.
The owner compares a later start, a narrower new assignment and suitable review cover. The decision keeps the consequence visible instead of assuming the new client will receive priority if the existing work expands.
This is scenario testing, not evidence of a precise failure probability. State the assumptions and use actual observations when available.
Review whether the protection helped
Check why reserve was used, what remained unused and whether commitments still failed. Persistent unused protection may justify adjustment; repeated overload may indicate unstable scope, poor estimates or an insufficient service boundary. Avoid removing all reserve after one quiet week.
Related guidance: practical process-mapping guide.
