Resource Capacity Planning: How to Match Demand to Capacity

What Is Resource Capacity Planning, and Why Do Plans Overcommit?

Resource capacity planning, sometimes called resource capacity management, is the practice of comparing committed work to the capacity actually available to deliver it, by skill and by period. The arithmetic is simple; the difficulty is that nominal headcount overstates capacity by a wide margin and approved projects understate real demand.

That is the whole problem in one sentence, and it explains the familiar standoff. The plan says the team has 130 person-weeks this quarter and the committed work needs 120, so the portfolio is comfortably inside capacity. The engineering manager says there is no room at all. Both are reading their own numbers correctly.

The plan counted people and called it capacity. The manager counted the hours those people will actually have after leave, the support rotation, the unplanned production work and the meetings, and then added the maintenance nobody put on the project list. The gap between those two numbers is where overcommitment is manufactured, every quarter, in good faith.

How Is Resource Planning Different From Capacity Planning?

Capacity planning asks how much work the organization can take on in a period, by skill. Resource planning asks who specifically does which piece of that work. Capacity planning comes first: there is no point assigning named people to work the organization cannot absorb.

What Is Your Real Available Capacity?

Project resource capacity planning starts from nominal headcount. Subtract, in this order. Each deduction is uncontroversial on its own; the total is what surprises people.

• Annual leave. Known in advance and still routinely omitted. Use the actual booked and accrued figures, not an average.
• Public holidays. Trivial to calculate, universally forgotten in quarterly plans.
• Support and on-call rotations. If one engineer is on support each week, that is a full person removed from project work for the whole period, not a small overhead on everyone.
• Unplanned and business-as-usual work. Production incidents, urgent fixes, requests that arrive directly. Measure it for one quarter rather than guessing; most teams are shocked by the share.
• Meetings, admin and ceremonies. Standing commitments that do not advance a project deliverable.
• Ramp time. A new hire is not a unit of capacity for their first several weeks, and the person onboarding them is partly unavailable too.
• Context switching. Someone split across three projects does not deliver three thirds. Either load people to one or two projects, or apply an explicit discount and say so.

The output is a usable-weeks figure per team or skill per period. It will look low. Teams will recognize it immediately, which is the point: a capacity number engineering disputes is a capacity number nobody will plan against. For the wider view, see resource management in enterprise PPM.

Why Is Demand Almost Always Understated?

Capacity gets deflated by optimism. Demand gets deflated by scope.

Most capacity plans count approved projects, because those are the things with a record. The work that does not appear includes: maintenance of existing systems, compliance and audit response, support escalations that need engineering, requests from executives that bypassed intake, technical debt that is slowing everything down, and the second half of last quarter’s projects that were declared done.

Each of these is real work consuming real capacity, and none of it is on the plan. The result is a portfolio that fits on paper and does not fit in a calendar. A disciplined project intake process fixes part of this by giving the unofficial requests a route, but only part: maintenance and BAU need a standing allocation rather than an intake form.

The practical fix is to reserve capacity before allocating any. Decide what share goes to run-the-business before the portfolio competes for what remains, and publish that share so it cannot be quietly borrowed from.

How Granular Should Capacity Planning in Project Management Be?

This decision quietly determines whether the plan is maintainable. Too coarse and the bottleneck is invisible; too fine and nobody keeps it current.

Three Planning Granularities Compared

For quarterly portfolio planning, plan by skill and resolve to named people only inside the next few weeks. The most common error is drawing skill categories too broadly: “engineering” as one pool hides that the three people who can touch the payments service are committed to four projects. If a category contains a sub-group everyone fights over, split it.

Which Levers Actually Close the Gap?

Five are available. They are not equally effective, and organizations reliably reach for them in the wrong order.

1. Cut commitments. The only lever that works immediately and reliably. Defer or stop projects until committed demand fits. Everything below this is a way of avoiding it, and most capacity crises are the accumulated cost of that avoidance. Pair it with a project prioritization matrix so the cut is made on evidence rather than volume.
2. Re-sequence. Same work, different order, so peaks in one skill do not collide. Effective when the constraint is a short spike and useless when demand exceeds capacity across the whole period.
3. Reduce scope. Deliver the valuable 60% of three projects rather than 100% of one. Works well, and requires a sponsor willing to define what the 60% is.
4. Add contractors. Slower than it looks. Onboarding consumes the capacity of the people you are trying to relieve, and the net gain in the first quarter is often close to zero. Useful for a known, separable workstream; poor as an emergency measure.
5. Hire. The right answer to a structural shortfall and the wrong answer to this quarter’s. Count the ramp honestly and assume no net contribution in the first period.

The reliable pattern: organizations try levers four and five first because they require no one to say no, then discover in month three that the shortfall is unchanged and the people who were going to help are busy onboarding. Lever one was available the whole time.

What Does a Resource Capacity Planning Example Look Like?

An illustration with stated assumptions, not a benchmark. One engineering team, ten people, a thirteen-week quarter.

Nominal 130, usable 64. Slightly under half, and every deduction is one a manager would call obvious.

Now the demand side. Four approved projects estimated at 58 person-weeks, which comfortably fits 130 and fits 64 with room to spare. Except the estimates carry no contingency, two of the four depend on the same two senior engineers, and the maintenance backlog was never counted. Add a 20% estimation buffer (70) and the picture changes: the quarter is at 109% of usable capacity before anything goes wrong.

The decision this forces is the useful output: cut roughly one project’s worth of commitment, or accept that one of the four will slip and choose now which one. Making that choice in week one is planning. Discovering it in week nine is reporting. Once a project does start to slip, its Project Health Score moves out of the on-track range and shows it.

When Is More Precision a Waste of Time?

Capacity planning attracts precision theater: plans accurate to half a day built on estimates that are comfortably plus or minus 40%. The decoration costs maintenance effort and buys no accuracy.

Useful tests for whether more precision is worth it:

• Would a different number change the decision? If the portfolio is at 140% of capacity, refining to 138% changes nothing. Cut first, refine later.
• Is the precision below the noise floor? Planning in half-days against estimates with a 40% error band is measuring with a micrometer and cutting with an axe.
• Who maintains it? Any granularity requiring daily updates from people who are not paid to update it will be abandoned, and an abandoned precise plan is worse than a maintained rough one.
• Does it survive one departure? If a single resignation invalidates the whole model, it was a staffing spreadsheet rather than a capacity plan.

Plan quarterly by skill at week granularity, and resolve to named people and days only for the next few weeks. For the longer-range view and the tooling question, workforce capacity planning tools covers the enterprise end, and demand vs capacity planning covers how the two disciplines fit together.

Where Does the Data for This Come From?

From the systems people already record work in, because a capacity model maintained by hand is stale before the quarter starts. Profit.co’s project portfolio management software keeps the capacity picture attached to live project records rather than in a parallel spreadsheet.

See how Profit.co shows committed demand against real capacity by skill. Book a demo.