Here is the quiet way most project portfolios fail. Leadership approves a set of projects that each look reasonable on its own. Nobody adds up the people those projects require. Three of them need the same senior engineer. Two need the only person who understands the legacy system. The plan is fine on paper and impossible in practice, and the first anyone hears of it is when deadlines start slipping for reasons no single project can explain.
Capacity planning prevents that. It is the discipline of comparing the demand your portfolio creates against the supply of people you actually have, and refusing to commit beyond it. It is unglamorous and it is the single most reliable predictor of whether a portfolio delivers.
Key takeaways
- Track capacity, not just demand. Over-commitment is invisible until you model both sides.
- Plan at the level of constrained roles and named specialists, not headcount averages.
- Leave slack. A portfolio planned to 100 percent utilization has no room for the work that always appears.
Demand versus supply, in plain terms
Demand is the total effort your approved and proposed projects require, ideally broken down by role and over time. Supply is the effort your people can realistically contribute, after you subtract holidays, support duties, meetings, and the operational work that never shows up in a project plan. Capacity planning is simply keeping demand below supply, role by role, period by period.
The reason this is hard is that demand is easy to see and supply is easy to overestimate. Every project advocates for its own resourcing. Nobody advocates for the truth that a person is already 130 percent committed.
Plan at the level of the constraint
Averages lie. A portfolio can look 80 percent utilized in total while the two people who matter most are buried. Effective capacity planning happens at the level of constrained roles and named specialists: the architect everyone needs, the one data engineer, the product lead who is on every steering committee. Find the constraints and plan against them, because they, not your headcount average, set the true throughput of the portfolio.
Set utilization targets you can actually sustain
A common mistake is planning teams to full utilization. It feels efficient and it guarantees failure, because real work includes interruptions, rework, onboarding, and the unplanned requests that always arrive. Sustainable utilization for project work is usually well below 100 percent. The exact number depends on how much support and operational load the team carries, but planning to the brink leaves no room to absorb the inevitable, and the result is missed dates and burnout.
Connect capacity to prioritization
Capacity planning and prioritization are two halves of the same decision. A ranked list of projects is only meaningful once you draw the line at where capacity runs out. Everything above the line is funded and staffed. Everything below it waits, no matter how appealing. If you have not yet built that ranking, start with how to prioritize a project portfolio, then bring the result here and test it against real supply.
Make the capacity conversation routine
Capacity is not a one-time model. People leave, projects slip, and new demand arrives constantly. The portfolios that stay healthy review demand versus supply on the same cadence they review priorities, usually inside the same governance forum. That way, the moment a new project would push a constrained role over the line, the tradeoff is explicit: something else has to move. For how to run that forum, see project portfolio governance, and for how a PMO turns these signals into decisions, start with the PMO overview.