Agency resource planning: forecasting capacity across roles and projects

Agency resource planning is the quarterly job of deciding which roles will run short before the weeks arrive when you can still do something about it. A weekly capacity check tells you your motion designer is overbooked this Thursday. Planning tells you that you will need roughly one extra motion designer for seven weeks next quarter, which is a different decision with a different lead time.

The two do not replace each other. If you already run the weekly committed-versus-bookable check described in creative agency resource management, keep it. This is the layer above it.

Planning answers questions the weekly check cannot

Once a quarter, per role, you need answers to four things:

• How many hours of each role does the pipeline imply over the next 13 weeks?
• Where is the gap concentrated — three bad weeks, or a steady shortfall?
• Can it be rebalanced internally, or does it need money?
• If it needs money, when does the person start, given ramp time?

Everything below is arithmetic on top of those four questions. The numbers in the examples are illustrative assumptions, not benchmarks — replace them with your own before you use them on anyone.

Build the plan on billable capacity, not headcount

The most common planning error is treating a 40-hour week as 40 hours of production. It is not. Standups, client calls, internal review passes, admin and the unbillable part of pipeline work consume them.

Then multiply by the number of people in that role and by weeks. Three editors at 30 hours across 13 weeks is 1,170 planning hours for the quarter. That single number is what the pipeline gets measured against.

Keep the non-production column honest by counting it once, the way you would explain it to a new hire — not the way you would explain it in a pitch.

Forecast committed hours, weighted by probability

Separate pipeline into three buckets and only plan with two of them:

• Signed. Contract or written confirmation. Plan with 100% of the hours.
• Probable. Briefed, scoped, waiting on a signature. Plan with 40–60% of the hours.
• Speculative. A conversation. Exclude it from the plan and track it separately.

A worked example for editors in one quarter:

• Signed projects: 780 editor hours.
• Probable projects: 520 hours at 60% probability = 312 weighted hours.
• Forecast: 1,092 hours.
• Planning capacity: 3 editors × 30 hours × 13 weeks = 1,170 hours.

That is 78 hours of headroom — about 6.7%. Thin, but covered. Now a probable project converts to signed and adds 300 hours. The forecast becomes 1,392 and the gap is 222 hours.

Two things follow. First, the gap is small enough that rebalancing may cover it. Second, 222 hours is not a headcount decision on its own: 390 hours is one full-time-equivalent for the quarter, so 222 hours is roughly 0.57 of a person.

That 222-hour figure is the size of the gap for the quarter. It says nothing yet about the shape — whether those hours are spread evenly across thirteen weeks or crammed into three in March. Answer the shape question separately by looking at the start weeks of the projects behind the hours.

Quarterly capacity estimator

Estimate how many full-time-equivalent production people one role's committed hours require across the quarter, before deciding whether to hire, contract or rebalance. It sizes the total gap for the period only, not how those hours are spread across the weeks or how they compare with other roles.

Required full-time equivalents equal committed production hours divided by the hours one person can produce in the period (weekly planning capacity multiplied by weeks).

Assumptions: Illustrative inputs, not measured benchmarks. Hours should be production hours only, excluding meetings, admin, review passes, holidays and contractor ramp-up. Run it per role: the result is the total gap for one role over the period, and it does not show whether the shortfall is spread evenly or concentrated in a few weeks — check project start weeks separately. Compare the result against your current bench to size the gap.

Worked example: Committed production hours in the quarter: 3,900 hours; Planning capacity per person per week: 30 hours/week; Weeks in the planning period: 13 weeks. Full-time equivalents needed: 10 people.

Then add ramp. A contractor who starts in week three of the quarter loses roughly two weeks of half-productivity, so 222 hours of shortfall becomes closer to 270 hours to backfill. Planning the gap without ramp time is how a quarter starts short and ends short.

The hire-or-contract trigger

Percentages beat gut feel here, because a 200-hour gap sounds alarming in a five-person team and trivial in a forty-person one. Measure the gap as a share of that role's quarterly planning capacity: (forecast hours − planning capacity) ÷ planning capacity × 100.

Two honest caveats. First, run this per role, never on a single overall number: a motion graphics shortage hides easily behind a producer surplus. Second, a hire has its own lead time — four to eight weeks to find and onboard someone in most creative roles — so the quarter you decide to hire is usually still a contractor quarter. Plan both.

Also check shape before size. A 300-hour gap spread evenly across 13 weeks is a different problem from the same 300 hours crammed into three weeks in March. The second one is often solved by moving one project's start date, which costs a client conversation instead of money.

Moving work between people without breaking commitments

Internal rebalancing is cheaper than hiring, but it is not free. Each move has a handoff cost and a client-facing risk. Four rules keep it from causing the slip you were trying to avoid:

1. One owner per deliverable at any moment. Two people half-owning an edit is how notes get lost.
2. Move at stage boundaries, not mid-revision. Handing over a file in the middle of round two means the new editor inherits context they cannot reconstruct from the notes.
3. Count the handoff cost. A 20-hour edit moved to someone unfamiliar with the client's format typically costs 2 to 4 hours of ramp. That is often still cheaper than a two-day slip — but budget it, do not assume it.
4. Do not move work off the critical path in its final week. If the deliverable has a client review scheduled, find capacity elsewhere.

Where a move changes a date, tell the client explicitly and early. Silent date changes are what turn a resource problem into a trust problem.

What to show leadership

Leadership does not need the spreadsheet. It needs one page, per role, with four numbers and one requested decision:

• Committed hours (signed).
• Weighted pipeline hours (probable × probability).
• Planning capacity.
• Gap in hours and as a percentage.

State the assumptions inside the page: the probability weights, the billable capacity per role, and the ramp allowance. A plan with hidden assumptions gets argued about instead of decided.

Say plainly that it is a forecast, not a commitment, and that you will re-run it monthly. A quarterly plan that is never updated becomes a document people work around.

A 90-day plan you can run on a board

You do not need dedicated resource planning software to run this. A board with six columns and one card per role works, whether that is a project board in your existing tool or a separate sheet. The columns are:

1. Scope — which projects are in this quarter's plan.
2. Forecast — committed and weighted hours per role.
3. Gap — hours and percentage per role.
4. Decision — rebalance, contract, or hire, with an owner and a date.
5. Staffed — the decision is actioned; named people or a signed contractor agreement.
6. Weekly check — linked to the live weekly capacity check rather than duplicated.

Put a card per role (four to six cards is realistic) and attach the biggest projects to the relevant cards with their hour estimate, start week and probability. If your tool supports board automations, the useful ones here are small: a notification when a card enters the decision column, and a nudge when a card passes its decision date. Those are the two points where plans quietly stall.

If the plan surfaces stalled or at-risk work as a side effect, the system for tracking stalled deliverables is a reasonable next layer. For the stages work moves through once it is staffed, see the brief-to-delivery workflow.

To start today: pick one role. Total its signed hours for the quarter, add its probable hours at 60%, and compare that forecast against the role's planning capacity. Compute the gap as a percentage — (forecast − capacity) ÷ capacity × 100. Under 15%, move a start date or shift a project week. Between 15% and 40%, contract for the quarter. Above 40% for two quarters running with signed work, plan a hire and contract for the current quarter anyway. Either way, you are making the call while the weeks are still ahead of you.