Don't Panic — Project Resource and Capacity Planning

Project resource and capacity planning is the awkward adult conversation where someone asks whether the people on the slide actually exist in the calendar. Demand is the work you intend to do. Supply is the productive capacity you can honestly field. Capacity is not headcount multiplied by optimism.

Before this discipline got a name, teams still made staffing guesses. They just made them inside Gantt charts that pretended every specialist could be in three meetings and two cutovers at once. Government standards and schedule assessors eventually said the quiet part aloud: if the resources are not there, re-plan the work. Do not keep the date and hope HR is a time machine.

Three ideas carry almost everything else.

First, use one unit. FTE or person-days let backend engineers, security reviewers, and lab time share a common language. Without that, portfolio math becomes vibes.

Second, load then resolve. Assign real capacity to activities, look for peaks above the line, then level or smooth. Leveling protects supply and may move the finish. Smoothing protects a hard date using float. If float is gone, you are no longer smoothing; you are choosing scope, sequence, or new supply with lead time.

Third, check the collisions across projects. Three plans at forty percent each can still overdraw one person in the same week. Local green is not portfolio green.

The surprise for many readers is that high utilization is not a gold star. It often means you spent the slack that unplanned work needed, multiplied multitasking, and then wondered why finished outcomes slowed while everyone looked busy. Critical-chain thinking makes the same point from another angle: the governing path includes resource contention, not only arrows between tasks.

If a spreadsheet full of names a year out comforts you more than a near-term named plan plus a far-term skill forecast, treat that comfort as a warning light.

What to read next: the Intro for the full map; the Cheatsheet when you need leveling versus smoothing in one glance; Practice for the arithmetic; Field Notes when you want the failure modes teams repeat; Landscape when you are choosing tools that show overallocation instead of hiding it.