Your agency has 60 free hours next week. A new project needs 20. On paper, the answer looks easy.
Then you discover that the project needs a senior developer, while most of the free time belongs to designers. The senior developer is already booked. Team-level capacity was accurate, but it did not answer the allocation question.
Resource allocation assigns suitable people to specific work at a time they can do it. A workable plan needs skills, availability, dependencies and room for the work that does not appear on a project timeline.
Start with each person’s actual availability
Begin with contracted working hours. Subtract leave and known unavailability, then account for internal responsibilities and existing project commitments. Use consistent definitions so an activity is not subtracted twice.
Do not assume every remaining hour is equally usable. Four scattered half-hours may be enough for reviews but poor conditions for a complex build. Look at the shape of the week as well as its total.
For the broader calculation, see capacity planning basics. Allocation takes that available capacity and makes a specific commitment.
Check the constraint before promising the work
Consider this illustrative weekly plan. Project capacity is what remains after leave and internal obligations. Spare hours are project capacity minus existing bookings.
| Person | Relevant skill | Project capacity | Already booked | Spare hours |
|---|---|---|---|---|
| Maya | Senior development | 30 | 28 | 2 |
| Leo | Design | 30 | 18 | 12 |
| Sam | Content | 28 | 14 | 14 |
The team has 28 spare hours, but a task requiring 12 senior development hours still does not fit. Maya has only two available hours. Moving it to Leo would require a real skills assessment and probably support, not just a calendar reassignment.
Your options are concrete: change the start date, move an existing commitment with agreement, narrow the work, or bring in suitable additional capacity. Pretending the team average solves the shortage only postpones the conflict.
Separate confirmed work from tentative demand
A proposal is not the same as a signed project. Show both when planning, but distinguish them clearly. Record tentative holds with a decision date and an owner who can confirm or release them.
Probability-weighted demand can help at portfolio level. A 50% chance of a 40-hour project contributes 20 hours to an expected-demand calculation. It does not mean the project will need half a person if it lands. Check the full scenario before promising a delivery date.
For overlapping opportunities, ask what happens if both are won. That conversation is more useful before signatures than after two clients have been given the same start date.
Allocate around dependencies and handoffs
An available designer cannot finish a page that depends on unapproved messaging. A developer may have capacity today but no approved design to build from.
Record the dependency beside the booking. Give uncertain inputs a decision point. If a client approval slips, decide whether to move the booking or use the time on work that can genuinely proceed.
Avoid leaving displaced work floating outside the plan. Every moved block needs a new date or an explicit unscheduled status, otherwise the same capacity gets sold twice.
Compare the plan with actual effort
Review planned and tracked time using the same person, project and period. If a task was booked for eight hours and consumed 14, investigate the cause before changing future allocations.
Was the estimate incomplete? Did scope change? Was the person interrupted by another client? Did the booking exclude review or coordination? Each explanation suggests a different correction.
Treat the comparison as planning feedback, not a personal efficiency score. Actual time includes the effects of the work environment and decisions made around the person doing the work.
Time Trakkr’s scheduling, leave-aware capacity view and planned-versus-tracked reporting can support this review. Use the feature overview to connect the schedule with delivery records rather than maintaining two disconnected versions of the week.
Common questions
Should we schedule every available hour?
Only if the plan genuinely accounts for all expected work and uncertainty. Choose an explicit allowance for interruptions and variable demand based on your own delivery history; there is no universal buffer that fits every agency.
How often should allocations change?
Review the near-term plan regularly and update it when a material dependency or commitment changes. Avoid constant reshuffling for minor requests: switching work has consequences even when the calendar looks tidy.


