Capacity Planning
Capacity planning compares the work an organisation has committed to against the people available to do it, over a defined horizon. Demand comes from project plans and resource plans; supply comes from headcount, contracts and realistic availability. The output is a picture of where demand exceeds supply and by how much.
What it contains
Capacity planning needs two datasets and one horizon. Demand is the sum of allocations across all committed work, broken down by role or skill and by time period. Supply is the people available in those roles, adjusted for leave, non-project duties, notice periods and the fact that nobody delivers a full working week to a single project. The horizon is usually two to four quarters — far enough to allow recruitment, near enough to be worth modelling.
The comparison is normally shown per role per month or per week: demand hours, supply hours, and the gap. A heatmap makes the peaks obvious, which is the presentation used in the capacity planner — over-allocation shown as a colour rather than a number you have to calculate.
Mature capacity views also separate committed demand from pipeline demand. Projects that are approved and staffed are a different class of claim from projects sitting in the intake list, and mixing them produces a shortage that may never materialise.
How it is used
The primary use is sequencing. When the portfolio wants three projects to start in the same quarter and the integration team can staff one, capacity planning is what turns that from a surprise in month three into a decision in month zero. The portfolio view supplies the list; the capacity view supplies the constraint.
The second use is lead time. If a shortage appears four months out and hiring takes three, the decision window is now. Capacity plans that only look one quarter ahead report shortages too late to act on.
The third is contracting. Where internal supply cannot meet demand, the gap becomes a statement of work, which links capacity planning directly to vendor selection and the budget that follows.
Where it goes wrong
The dominant failure is planning against theoretical availability. Supply calculated from headcount times working days overstates reality substantially once leave, support duties, interviews, line management and organisational overhead are counted. The plan then shows capacity that does not exist, and every project starts on time and finishes late.
The second is aggregating roles that are not interchangeable. "Twelve developers available" means nothing if the work needs the three who know the payments platform. Capacity is only meaningful at the level of skill that the work actually requires, which is usually more granular than the org chart.
The third is planning capacity without planning the peaks. Cutover weekends, UAT cycles and post-go-live support draw heavily on the same people, often simultaneously across projects. A steady-state capacity model will show those months as comfortable.
The fourth is a capacity plan maintained by the PMO and ignored by the line managers who control the people. The numbers are correct and nothing changes, because the plan has no route to a decision. That is a governance problem rather than a modelling one, and the guide to programme governance covers where those decision rights should sit.
Related terms
Resource plan is the demand side for a single project. Utilisation is the proportion of available capacity that is assigned; high utilisation reduces the slack that absorbs variation. Demand management is the intake process that generates the demand figures. Bottleneck is the role or team whose capacity constrains everything else. Portfolio management is where capacity constraints are supposed to change decisions.
Questions
What horizon should capacity planning cover?
Long enough to act on what it shows. If recruitment takes three months, a one-quarter horizon reports shortages after the window to fix them has closed.
Should capacity be planned in hours or days?
Either works. Days per week is easier to maintain and easier for line managers to sanity-check; hours suit organisations already tracking time that way.
How is unapproved work handled?
Keep pipeline demand separate from committed demand. Combining them produces shortages driven by projects that may never be approved.
Why do capacity plans overstate availability?
Because supply is usually calculated from headcount and working days without deducting leave, support work, line management and other non-project time.