Agency capacity planning: a worked example for a 25-person agency

Agency capacity planning means comparing the hours each role can actually spend on client work with the hours already sold or likely to be sold. Do it by role, by week, and with confirmed and pipeline work kept apart.

This guide walks through one month at a 25-person agency: how to work out capacity, how to read booked against forecast, and the decisions it leads to. All names, projects and rates are illustrative.

Planning in a spreadsheet? Download the free capacity planning template. It uses the same method as this example.

Capacity planning in one line

Capacity planning answers “can we take this work?” Resource allocation answers “who does it?”

Capacity planning looks at supply and demand in aggregate, by role. It tells you design is 90 hours short in the week of 16 November, not which designer moves. Once you know the gap, use the resource allocation template to plan person by person.

Booked vs forecast: the two numbers you need

Every capacity plan needs two demand lines:

• Booked (confirmed): signed retainers and projects. The work is happening.
• Forecast: booked work plus pipeline work weighted by its chance of winning. A £40k pitch at 60% counts as 60% of its hours.

Booked tells you whether you can deliver what you've sold. Forecast tells you whether you can sell more, and when you'll need help. Look at only one of them and you either turn down work you could have taken or win work you can't staff.

Step 1: work out available hours by role

Start with contracted hours. Take off time off. Then apply the share of time that never goes on client work: management, new business, internal meetings, training.

Available hours = (contracted hours − days off × hours per day) × (1 − non-project %)

Our example agency has 25 people on a 37.5-hour week (7.5 hours a day). November 2026 has 21 working days, so each person is contracted for 157.5 hours.

Two things stand out before any project goes in:

1. Only two-thirds of contracted time is sellable. 2,626.5 of 3,937.5 hours, or 67%. That matches the 61–72% blended utilisation a 25-person agency can realistically plan for (see the benchmark on our agency resource management page).
2. Non-project shares differ by role. Directors typically bill 30–40% of their time, delivery staff 70–80% (Alto Accounting, citing BenchPress and Synergist). One agency-wide figure would overstate director capacity and understate design. To set your own figures, use the billable utilisation calculator.

Step 2: list the work, confirmed and tentative

The agency has four confirmed jobs and three in the pipeline for November.

Hours are for November only. Set the chance of winning from your own history, not optimism. If you don't track win rates, use 25% for a cold pitch, 50–60% for a shortlist and 90% for an extension.

Step 3: compare booked and forecast against capacity

At agency level, 78% booked and 97% forecast looks healthy: busy, not broken. That is the problem with monthly, agency-wide numbers. They hide where it breaks.

Step 4: go weekly where a role is over 100%

Design is at 104% forecast for the month. That sounds like a little overtime. Split by week, it isn't.

Available hours change week to week because time off isn't spread evenly: 4 of the 9 design days off fall in the week of 23 November.

What the weekly view shows:

• Week 1 has 74 spare design hours. The monthly figure hid them.
• The pressure is all in the second half of the month, where the rebrand pitch and the Christmas extension both land. If both are won, design is 178 hours (about 24 days) short between 16 and 30 November.
• The weighted forecast is a planning number, not a staffing number. No pitch arrives at 60%. It arrives at 0% or 100%. Use weighting to see the trend, then plan staffing for the real outcomes.

Step 5: plan for outcomes, not averages

Development shows this most clearly. The weighted forecast is exactly 100%, which looks “full but fine”. In reality there are two possible Novembers:

Neither of those is 100%. Plan for each one separately.

Step 6: make the calls

This is what the agency's resource manager decides on 30 October, before any pitch result is known:

1. Pull work forward into the spare week. Move 40 hours of website rebuild design from the weeks of 16 and 23 November into the week of 2 November, where there are 74 hours spare. This costs nothing and cuts every scenario's gap by 40 hours.
2. Put a provisional hold on a freelance designer. If the rebrand and extension are both won, design is still 138 hours (about 18 days) short after the move. Book the freelancer provisionally now and confirm when the pitch result lands. Freelance cover changes the margin: 18 days at £550 a day costs £9,900, against £7,560 for the same days internally at £420. That's £2,340 less margin, but the work gets delivered on time.
3. Put the start date into the app MVP proposal. If the app MVP is won for November, development is 180 hours short. Rather than hiring for a 30% chance, propose a December kick-off in the pitch. A start date you agree before signing costs nothing. One you ask for after winning costs goodwill.
4. Watch, don't act, on strategy, copy and account management. Each reaches 108% only if every pitch lands, and that has roughly a 16% chance (60% × 30% × 90%). Moving deadlines inside the month covers it.
5. Keep pitch work out of director delivery time. Directors are at 32% booked, inside the 30–40% benchmark. The pitches themselves come out of their 65% non-project share, so adding pitch hours as “billable” would make director capacity look worse than it is.

The hiring test: hire when a role is short for many weeks in a row, not for a one-week peak. Nothing in this November justifies a hire. A design shortfall that repeats in December and January would.

A monthly capacity planning routine

1. Weekly, 15 minutes: update time off, confirmed bookings and pitch probabilities.
2. Monthly: look 4, 8 and 12 weeks ahead by role. Flag any role over 100% forecast or under 60% booked. Both cost money: one in freelancers and missed deadlines, the other in salary you pay for and don't sell (Alto Accounting puts the warning line below 60%).
3. At every pitch decision: check what winning would do to each role's weekly load before you say yes.
4. Quarterly: compare forecast utilisation with actual. If your forecast is regularly 10 points above actual, your non-project shares or win rates are too optimistic.

Common agency capacity planning mistakes

• Planning at agency level. 97% overall hid a design team at 125% in two weeks.
• Planning by month. The spare week at the start of November disappears in a monthly total.
• Ignoring pipeline. Booked-only plans say yes to pitches the team can't staff.
• Treating weighted forecasts as real. A 30% pitch never needs 30% of a developer.
• One utilisation target for everyone. Directors at 80% billable is a plan nobody will hit.
• Forgetting time off. Nine design days off in a month is 54 available hours gone after non-project time. Leave clusters around holidays.

When a spreadsheet stops working

The example above is right on the day it's filled in. Then someone books leave, a pitch is won, the rebuild slips a week, and nobody re-enters twelve weeks of numbers.

Kavaro is resource management software for agencies and consultancies, and it keeps this picture live. The Schedule shows each person's bookings over the coming weeks, with tentative work kept apart from confirmed work and open roles as their own rows. You can ask the agent “can we take the charity rebrand from 16 November?” and see what it does to each role, and to margin, before you approve anything.

See next month's capacity by role, live: start a free 30-day trial. No credit card needed.

Or read how it works for agencies: agency resource management software.

Questions people ask

What is agency capacity planning?

Working out how many hours each role can spend on client work over the coming weeks, and comparing that with work already booked and work in the pipeline, so you know before you commit whether you can deliver it and when you'll need freelancers or new hires.

How do you calculate capacity for an agency?

For each person: (contracted hours − days off × hours per day) × (1 − non-project share). Add people up by role, then compare each role's total with confirmed hours and weighted pipeline hours, week by week.

What's a good capacity utilisation for an agency?

Plan for about 70–80% billable for delivery roles and 30–40% for directors. A 25-person agency typically blends to 61–72%. A delivery team that stays above 85% is a burnout risk; below 60% means you're paying for time you don't sell.

How should pipeline work be counted?

Weight each pitch's hours by its chance of winning to see the trend, then check the won and lost scenarios for any pitch big enough to push a role over 100%, because the work will arrive at full size or not at all.

How far ahead should an agency plan capacity?

Detailed planning 4 weeks ahead, by role and week, and a rougher view 8–12 weeks ahead to spot hiring needs and pitch clashes.

What's the difference between capacity planning and resource planning?

Capacity planning compares supply and demand by role. Resource planning, or allocation, names who works on which project each week. Do capacity first: it tells you whether a plan is possible before you assign people.

Related

• Capacity planning template (Excel and Google Sheets): the method in this guide, ready to fill in
• Agency resource planning template: a quarter-ahead plan with pitches weighted by chance of winning
• Resource allocation template: person × project × week
• Billable utilisation calculator: set realistic targets by role
• How to calculate utilisation rate
• Agency resource management software