Resourcing and capacity planning: the complete guide

In this guide

1. What resourcing and capacity planning covers
2. Why agencies get this wrong
3. The cadence that keeps it working
4. Reading utilisation without over- or under-booking
5. Spreadsheets vs dedicated software
6. Frequently asked questions

Every agency runs into the same two questions on a loop: who is free to take this project, and are we about to run out of people. Resourcing and capacity planning is the discipline that answers both, and most firms answer them badly not because the questions are hard, but because they treat allocation, capacity, and forecasting as one blurry problem instead of three connected ones.

This is the complete guide to that discipline: what each of the three pieces actually does, why agencies default to reactive firefighting instead of planning ahead, the cadence that keeps a bench healthy, and how to tell whether a spreadsheet is still good enough or whether it is quietly costing you. Wherever a topic has its own deeper post, this guide gives you the short version and links to it.

What resourcing and capacity planning covers

Resourcing and capacity planning covers three related but distinct disciplines: allocation, capacity, and forecasting, each answering a different question on a different time horizon.

Allocation is the act of assigning named people to specific project work on specific dates. It answers "who is doing this task, and when." Our resource allocation guide covers the methods agencies use and how to avoid over- and under-booking any one person.

Capacity is whether the team has enough available hours to take on a piece of work before you commit to it. It answers "do we have room for this," usually over the next few weeks. Resource capacity planning covers how to calculate available capacity and the weekly habit that keeps it accurate.

Forecasting is projecting staffing needs further out, based on the pipeline of work that is not confirmed yet. It answers "what will we need in six or eight weeks," which is a very different question from "what do we need today." Resource forecasting vs capacity planning walks through why the two horizons need separate processes even though they feed each other.

The three disciplines, and the question each one answers

The three build on each other in one direction. Forecasting tells you roughly what capacity you will need. Capacity planning confirms whether you actually have it once a piece of work is real. Allocation is where that confirmed capacity turns into named people on named tasks. Treat them as one undifferentiated activity and you end up doing all three badly at once: allocating people before you've checked capacity, or checking capacity without any sense of what is coming next quarter.

Confusing the three is where a lot of agencies go wrong in practice, not because the definitions are hard, but because one tool or one spreadsheet gets stretched to answer all three questions at once. A tab built to track this week's assignments gets pressed into service for a six-week staffing forecast, and it does neither job well: it's too detailed to reason about the medium term, and by the time anyone updates it for the forecast, this week's allocation view is already stale. Keeping the three questions separate, even if they live in the same tool, is what makes each one answerable in the time it deserves.

Why agencies get this wrong

Agencies get resourcing wrong by running it reactively, deciding who works on what only after a deadline is already close, instead of running it as a standing weekly process. Reactive resourcing feels efficient because nobody spends time on planning meetings, but it produces the exact pattern it's trying to avoid: some people buried in overtime while others sit half-booked, discovered only when someone finally checks.

The reactive pattern usually starts the same way. A new project lands, and whoever is available gets pulled onto it, because checking real availability takes longer than asking around. That person may already be stretched thin on two other accounts, but nobody had a current view of that, so the assignment goes ahead anyway. A few weeks later the same person is behind on all three projects, and the fix is another reactive assignment, borrowing someone from a different team who is now also overloaded. Each fix creates the next fire.

The reason this keeps happening even at agencies that know better is that resourcing data goes stale fast. A capacity view built at the start of the month is wrong by week two, once new work has landed, someone has taken unplanned leave, and a project has run longer than scoped. Treating capacity planning as a one-off exercise done at the start of a quarter is the same mistake as treating a budget as accurate forever, once it's set. It needs a refresh cadence, not a one-time build.

Here's what that looks like in practice. A five-person delivery team is running at what looks like healthy utilisation on paper, three projects, everyone nominally booked. One senior consultant is actually carrying pieces of all three, because they're the only one who's worked with that client type before, and nobody totalled up their hours across projects when the third one got assigned. Four weeks in, that person is working evenings to keep all three on track, two of the projects are quietly behind, and the first anyone hears about it is when the consultant flags they're burned out. The data to catch this existed the whole time, spread across three project plans; nobody had it in one place at the point the third project got assigned.

The fix is not more planning meetings. It is checking capacity before committing to new work rather than after, and doing that check often enough that the numbers are still true when you act on them. That's a cadence problem, which is the next section.

The cadence that keeps it working

The cadence that keeps resourcing working is a short weekly check on current allocation and capacity, paired with a less frequent monthly look at the pipeline for forecasting. Different questions need different refresh rates, and forcing them into the same meeting is why resourcing reviews either run too long or skip the parts that matter.

Weekly: allocation and capacity. Once a week, confirm who is actually booked on what for the coming two to three weeks, and flag anyone over 100 percent or sitting well under target. This catches over-booking before it becomes a missed deadline and under-booking before a week of billable hours is lost to nobody noticing someone was free. Fifteen minutes is usually enough once the data is current, because the review is confirming a live plan, not rebuilding one from scratch.

Monthly: forecasting. Once a month, look further out: which deals in the pipeline are likely to close, what staffing they would need, and whether current capacity can absorb that without new hires or reshuffling. This is a slower, more speculative conversation, and it belongs in its own session because mixing "what's happening this week" with "what might happen in six weeks" tends to crowd out the near-term fires that actually need attention that day.

The mistake worth naming directly: many agencies run this backwards, spending real time on quarterly staffing projections while the weekly allocation view is stale or nonexistent. The forecast tells you what you might need. It's the weekly check that stops someone from being double-booked next Tuesday, and that near-term failure is the one clients actually notice.

Reading utilisation without over- or under-booking

Reading utilisation well means checking it against a target range, not against 100 percent, because full utilisation almost always means someone is over-committed rather than perfectly efficient. Utilisation is the share of a person's available hours booked to billable work, and a number close to 100 percent looks great on a dashboard right up until the first unplanned absence or scope change, at which point there is no slack left to absorb it.

Most agencies do better targeting 75 to 85 percent utilisation for people doing client delivery work, leaving room for internal meetings, business development, and the inevitable day where a project runs long. The right target varies by role: a senior consultant carrying account and delivery responsibilities needs more non-billable room than a mid-level specialist executing defined tasks, so a single company-wide target usually undershoots for some roles and overshoots for others.

Over-booking is the more common failure and the more expensive one. It shows up as a person allocated past 100 percent across their active projects, which on paper looks like strong demand and in practice means at least one of those projects is quietly slipping. Under-booking is the quieter failure: a person sitting at 50 percent utilisation for two weeks straight because nobody assigned them work, which is lost revenue that never shows up as an incident the way a missed deadline does. Both failures are visible in the same data, which is exactly why the weekly cadence above needs to check current allocation against target range, not just glance at whether anyone is on fire this week.

A useful habit is to sort the team by allocation percentage before every weekly review, rather than scanning project by project. Project-by-project views hide the problem, because a person can look fine on each individual project and still be over 120 percent once their hours are added up across all three. Sorted by person, the over-booked and under-booked cases surface immediately, without anyone having to do the addition by hand.

Spreadsheets vs dedicated software

Spreadsheets can run resourcing for a small team, but they stop scaling once headcount and concurrent projects grow past what one person can hold in their head. A shared spreadsheet works fine for eight people on three active projects, because the person building it can just remember who's on what. It stops working once that grows to thirty people across a dozen engagements, because keeping allocation, capacity, and the pipeline forecast all current in the same file becomes a full-time reconciliation job that nobody actually has time for.

The tell that a firm has outgrown its spreadsheet isn't headcount on its own, it's how often the resourcing view is wrong when someone actually checks it against reality. If the weekly review keeps surfacing "wait, I thought so-and-so was free" or "this project isn't in here at all," the spreadsheet has already fallen behind and the team is resourcing off memory with a document as a rough backup. Our capacity planning software guide covers the specific features that separate a tool worth buying from a more elaborate spreadsheet, and consulting firm management software covers what to look for when utilisation and staffing sit at the center of the business model rather than being a side concern.

Dedicated software earns its place by keeping allocation, capacity, and the pipeline view connected to the same live data instead of three documents someone has to update separately. When a project's timeline shifts, capacity for the people on it updates automatically instead of waiting for the next manual edit. That connection is the same idea behind a PSA platform more broadly: resourcing data that's wired into the rest of the system, not a spreadsheet that has to be kept in sync by hand. See how Pike handles resourcing as part of the same connected system as projects and billing.

Frequently asked questions

See how Pike works for your team

Pike keeps allocation, capacity, and the project timeline connected, so booking someone to a task updates their capacity everywhere else in the system instead of living in a separate file someone has to update by hand. Whether you're checking who's free this week or forecasting what the pipeline needs next month, it's the same live data behind both views.

Book a demo at cal.com/usepike/demo to see how resourcing would run for your team, or check pricing to see what plan fits.