Where a capacity matrix answers "who is on what right now", the forecast answers "who will be free, and when". The per-person, per-week arithmetic is simple: free hours = weekly baseline (40) − allocated project hours − approved time off. Laid out over the coming weeks it becomes the operating map of a services business — what sales can promise, what staffing can commit, where the bench will appear.
What the forecast is for:
- Staffing without guessing — the next project starts from a list of who has hours in week 3, not from memory.
- Seeing roll-offs early. An allocation dropping to zero six weeks out is a sales-and-staffing conversation today; discovered in week six, it is a bench write-off.
- Catching overbooked weeks — a person summing past their baseline in a future week is a delivery risk that costs nothing to fix now.
- Making time off real. Approved vacations reduce available hours in exactly the weeks they cover — a forecast that ignores them overbooks the summer.
The useful horizon is 8–12 weeks: long enough to act on what it shows, short enough that the plan is still honest.
In Helia, the capacity forecast projects each person week by week — planned assignment hours minus approved time off against a 40-hour baseline, 12 weeks by default — with free hours and overbooked flags per cell; and when a billable assignment nears its end date, a reminder pings the project's delivery manager to either extend it or confirm the roll-off.