At ten people, you can plan capacity in your head. At sixty, you cannot, and the moment you try, projects start colliding. Resource and capacity planning for a growing architecture firm is the discipline of knowing who is available, who is overloaded, and whether you can actually take on the next project, across a team too large to hold in memory. Get it wrong, and you either burn out your best people or leave capacity idle while you turn down work.
Midsize firms hit a specific wall: the informal coordination that worked when everyone sat in one room breaks down, but the firm is not yet running on rigorous systems. That gap is where good people get double-booked and projects quietly slip.
The three questions capacity planning must answer
Strip away the complexity and capacity planning exists to answer a small number of questions clearly. If your system cannot answer these, it is not doing its job.
- Who is overloaded right now, and what can we take off their plate
- Who has room this week that we are not using
- Can we say yes to the new project without breaking a current one
- Which teams are consistently the bottleneck as we grow
Utilization is a signal, not a scoreboard
Utilization, the share of someone's time spent on billable project work, is the classic midsize-firm metric. It is genuinely useful for understanding whether the firm is healthy. It is also dangerous when it becomes a stick. Pushed as a personal target, it drives people to log hours defensively and hide slack, which corrupts the very data you need. Read utilization at the team and firm level to spot structural problems, not at the individual level to rank people.
Capacity data you use to police individuals is capacity data you can no longer trust.
The spreadsheet stops scaling around here
Almost every growing firm runs capacity on a spreadsheet, and almost every one outgrows it. The spreadsheet cannot keep up because it is a snapshot maintained by one person, and reality changes faster than they can update it. By the time the grid is accurate, it is out of date. At 50 to 100 people, capacity has to come from the actual work, not from a parallel document someone hand-maintains.
Connect capacity to real tasks
The way out of the spreadsheet is to derive capacity from what people are actually assigned. When every task has an owner and lives on a shared board, load becomes something you can see rather than something you estimate. You can look at a person or a team and see the real queue of work in front of them. That is far more honest than a planning grid, because it reflects commitments that already exist instead of intentions someone typed in a cell.
Plan the pipeline, not just today
Capacity is not only about this week. A growing firm has to look at the pipeline: the projects about to start, the phases about to ramp up, the people about to free up. The teams that plan well look a few weeks ahead and see the crunch coming, so they can rebalance before it becomes a crisis. That requires a shared view of both current load and what is on the horizon, visible to everyone who staffs projects.
Make it a rhythm, not a fire drill
Capacity planning fails when it only happens in emergencies. The firms that handle growth well build a short, regular cadence, a weekly look at load and the pipeline, so rebalancing is routine instead of reactive. It does not need to be a long meeting. It needs to be consistent and grounded in real data.
The foundation that makes it possible
You cannot plan capacity you cannot see. Spreadbox gives a growing firm that visibility: Kanban task boards and a team queue mean every assignment has an owner and a place, so load across a large team is something you can look at rather than guess. Project status tracking shows which phases are ramping up across the pipeline, per-task time tracking and a day clock give you honest data on where hours actually go, and team chat keeps staffing conversations in one place. Spreadie, the AI assistant, can summarize where things stand so your weekly capacity review is a quick read, not a scramble. Instead of a spreadsheet that is always slightly wrong, you plan from the real state of the work.