Capacity planning

Capacity planning looks like an operations exercise and pays off in sales. The moment it matters is when somebody asks whether the agency can take on a new account starting next month.

Most agencies answer that optimistically, and not out of dishonesty. The honest answer costs revenue now against a quality problem later, the pipeline has been slow, and the team has coped before. Every incentive points at yes.

The cost of the wrong yes is paid by existing clients, in slipped deadlines and thinner attention, and it is paid in a currency nobody attributes back to the decision. Three months later an account churns and the post-mortem does not mention the January pitch.

A capacity plan does not make the decision, it makes it explicit. Taking this on means two existing accounts get less in February is a sentence somebody can accept deliberately, hire against, or decline. Without the plan, the same trade happens and nobody chooses it.

A forecast that counts every open proposal as work produces a plan showing the agency at capacity when half of it will not close. Used to justify a hire, that is expensive. Used to decline a pitch, it is worse.

The correction is to weight by stage, and the weights do not need to be sophisticated. A proposal sent is not a proposal verbally agreed, and a verbal agreement is not a signature. Three bands applied consistently beat a precise model applied inconsistently.

The number worth knowing is the agency own historical close rate by stage, which most agencies could calculate and have not. It is the difference between a weighted forecast and a guess with percentages on it.

Two views are then worth holding at once: committed work only, and committed plus weighted pipeline. The gap between them is the decision space, and it is where hiring and pitching decisions actually get made.

Team-average capacity is almost always reassuring and almost always wrong, because agency work does not distribute evenly across people. It concentrates on whoever holds a specific skill.

An agency with spare hours everywhere except its one paid media lead has no spare capacity for anything involving paid media, which is likely most of what it sells. The average says there is room; the constraint says there is not, and the constraint is what decides whether the work ships.

So the plan has to be per skill, or at least per constrained skill. Most agencies have two or three: the person who does paid media, the person who does development, the person clients ask for by name. Those three schedules decide the agency capacity, and the rest of the team is downstream of them.

It also identifies the highest-value hire, which is frequently not the one being discussed. An agency planning to hire another generalist when the constraint is a single specialist will add cost without adding capacity.

Three things are consistently missing from agency capacity plans, and all three are knowable in advance.

Onboarding is the first. A new client consumes far more time in month one than its retainer implies: audits, access, tracking, strategy, learning the business. Planning a new account at its steady-state hours guarantees an overloaded first month, every time.

Absence is the second. Holiday, public holidays, notice periods and parental leave are all known weeks or months ahead and are routinely omitted, which is how a plan that worked in principle collapses in August.

Non-billable obligations are the third. Pitching, internal work, training and management are real hours that a plan built on billable time pretends do not exist. An agency planning at a hundred per cent of available hours has planned for a team that never pitches for anything.

Adding those three turns a plan from a description of the ideal month into a description of a real one, which is the only version worth making decisions against.