Resource Planning for Agencies: The Forecast-First Approach

Resource planning for agencies means checking the work coming through the pipeline against who will be available to do it, eight to twelve weeks out, while there is still time to adjust. The forecast-first approach starts with five questions about what the firm expects to sell and who it expects to have, then checks the plan against actuals every two weeks.

If you have ever sat in a Monday resourcing meeting and watched three project managers argue over the same senior designer for next Tuesday, you have seen what most resource planning advice gets wrong. Tuesday is where the argument becomes visible, not where it starts. The problem began four weeks earlier, when the work coming through the pipeline was never checked against who would actually be available to do it.

That is resource planning. Most of the guides you can find online describe the Tuesday argument. This one is about how the argument got there.

What is resource planning?

Resource planning is the process of matching incoming work to the people, time, and budget available to do it, looking far enough ahead that you can adjust before the matchup becomes a fire drill.

Definitions usually stop at the matching part. The “looking far enough ahead” part is where agencies feel the difference. Describing who’s working on what this week is a schedule. Describing who you’ll need, who you’ll have, and where those two lines diverge twelve weeks out is something else.

Why resource planning breaks down at agencies

When planning goes sideways, the first instinct is to blame the tool. Spreadsheets are too brittle. PM tools are too project-focused. CRMs do not talk to the time tracker. All of that is true. None of it is the actual problem.

Agencies break down on resource planning for a quieter reason. They have data in three places. Pipeline lives in the CRM. Capacity lives in the resourcing tool or, more accurately, in a spreadsheet that one person updates on Fridays. Time tracking lives somewhere else. Each system answers a different question. None of them connects the three questions that matter when leaders are making decisions about the next quarter.

Those questions: what work is most likely to come in. Who is most likely to be available when it does. What it costs the firm if those two lines don’t match.

A timesheet works as a receipt for reality. Pipeline forecasts work as a probability map. Capacity plans work as a hypothesis about how the next ninety days will play out. Resource planning is what you do when you put those three things in the same conversation. Most agency stacks weren’t built to.

The forecast-first approach: five questions to start with

Plenty of resource planning guides give you a process. Here is one too, but flipped. Instead of starting with the work and asking who can do it, start with the firm and ask what it is forecasting.

1. What is the firm planning to sell over the next ninety days? This is a pipeline question, not a sales question. Weighted pipeline, with realistic win probabilities, by service offering and projected start date. If your CRM cannot produce this without an hour of cleanup, that is a finding, not a flaw in this step.

2. What capacity will the firm actually have during that window? Subtract the obvious. PTO. Holidays. Internal projects. Then subtract the less obvious. Recurring meetings. Admin work that does not get billed. Onboarding load for new hires. What remains is the capacity you can plan against.

3. Where does the line of forecasted demand cross the line of forecasted capacity? This is the only question that matters, and it is the one that almost never gets asked directly. If demand exceeds capacity, you have a hiring or subcontracting decision coming. If capacity exceeds demand, you have a sales pressure problem coming. Either way, you need eight weeks of lead time to act on the answer.

4. What is your confidence in those two lines? Pipeline forecasts drift, PTO requests land, and scope changes hit mid-engagement. The forecast is a working hypothesis, not a contract. Knowing where the confidence is low tells you where to look more often.

5. What will you do when the lines diverge? This is the part that gets skipped. Most resource plans assume the plan will hold. Good ones plan for the reconciliation loop, which is where the work actually gets done.

If those five questions sound less operational than what you’d see in a typical resource planning guide, that’s on purpose. Operational steps get straightforward once these five are answered. Skip one, and they get hard fast.

A walkthrough: planning resources at a 60-person agency

Take a sixty-person creative agency with eight active client retainers and a small project pipeline. One senior designer who is everyone’s favorite. The COO runs a weekly resourcing meeting where the same conversation has happened for the last three months.

It sounds like this. “Can we get Maya on the rebrand?” “She is on the retail launch through November.” “What about December?” “December looks fine.” A week later, the launch slips two weeks. Maya is still on the launch in December, and the rebrand needs her too. Two months ago, none of this was visible.

A forecast-first approach would have caught it at week two of October. The tool would not have warned anyone. Asking “where does pipeline demand cross capacity over the next ninety days” would have surfaced that Maya was the single point of failure for two engagements with overlapping risk profiles. Finding another senior designer in December is too late. Making that hiring or subcontracting decision in October, while there is still time, changes the outcome.

This is not a hypothetical. We see patterns like this across the agencies we work with, and a scheduling tool almost never catches them, because scheduling tools answer last week’s question, not next quarter’s.

What to do when the plan and reality diverge

Plans drift. This is the part everyone admits and then writes a guide that pretends otherwise. A resource plan’s job is to stay useful when it is wrong, not to be right.

Useful in this case means three things.

You can see the divergence quickly. If your plan said Maya would be 70% billable in November and she ended up at 95%, you should know that by week one of December at the latest, not by quarter end. Real-time visibility into plan versus actual is the only way the loop closes.

You know what triggered it. Was it a scope change? A slipped engagement? A new client we didn’t expect to win? Each one suggests a different correction.

Adjustments don’t require redoing the plan from scratch. Plans that break every time something moves get abandoned. Good plans absorb small changes and only surface big ones.

Mechanics like this can live in different tools. Connective tissue is the harder part. Linking pipeline to capacity to actuals is the part most agency stacks are missing.

Tools that support resource planning

Compare resource planning software online and most of it does one of two things: it schedules, or it tracks. A smaller set forecasts. Fewer still connect forecasting to actuals.

Here is where the major options actually stand, without pretending they all do the same thing.

Float and Runn used to be a clean week-by-week assignment view and not much else. That has changed. Float’s Pro plan now tracks project budget burn, margin, and estimates versus actuals. Runn shows planned versus actual hours on every plan, including its cheapest tier. Both are reporting on a project you already scheduled inside their tool. Neither one pulls in a pipeline sitting in your CRM. “Will next quarter’s demand outrun the capacity I have” is still not a question either one answers. You get a sharper read on how an assigned project is tracking. You do not get a forecast.

Productive and Scoro dropped the all-or-nothing pitch too. Productive now sells itself on connecting to what you already run, with built-in syncs to HR platforms like BambooHR and accounting tools like QuickBooks and Xero. Scoro sells its platform as apps you add one at a time, priced separately, instead of one big rollout. That is a genuine change from the old pitch. But the destination has not moved. Once time, resourcing, and invoicing are running through their apps, that data lives in their system, whether it arrived in one migration or five smaller ones.

Monday and Asana are work management platforms with resource views layered on. They work well if your team already lives in them. They are not built specifically for resource planning, but for many agencies the gap is acceptable.

Spreadsheets still drive the planning layer in a surprising number of well-run agencies. Flexible, free, and quick when the question is not yet stable. Fragility appears under volatility. If your spreadsheet has more than four tabs and you have ever lost an afternoon to a broken formula, you have outgrown the spreadsheet.

Parallax adds a planning layer to the tools you already have. Instead of asking you to migrate your project management or time tracking, it connects pipeline, capacity, and actuals so the divergence question can actually be answered. The trade is that you keep your existing stack and add a planning lens over it.

All of this matters because no single tool wins this conversation. “Which tool is best” is a question with no useful answer. A better question: which gap in your current stack is hurting decisions, and what addresses that gap with the least disruption.

What good resource planning looks like in practice

A useful benchmark, since most posts on this topic skip it. Across the agencies we have worked with, the ones that have resource planning working well tend to look like this.

Capacity gets forecast at least eight weeks out, by role, accurately enough that senior leadership can answer “are we hiring or not?” without scheduling a meeting.

Weekly resourcing meetings end with decisions, not with “we’ll figure it out.” If a meeting ends without changes to the plan or the pipeline, it was a status update, not a planning session.

Plan-to-actuals reconciliation happens every two weeks. The reconciliation does not need to be precise. It needs to surface the gaps that matter.

Forecasts stay vague where they should be vague. A capacity plan that assigns specific names to projects ninety days out is fiction. A capacity plan that assigns roles and skill levels ninety days out is forecastable.

None of this requires new software. It requires the connective tissue between the systems already in place.

Common mistakes (and what to do instead)

A few patterns repeat over and over.

Treating the resource plan as a project plan. Project plans answer what gets done by when. Resource plans answer whether the firm has the capacity to do the work in the pipeline. They overlap. They aren’t the same thing. Confusing them produces detailed schedules that fall apart under scrutiny.

Planning at the individual level too far out. Naming Maya in November is fine. Naming her in February is fiction. Further out, plan by role and skill, not by person.

Updating the plan only when it breaks. Plans touched only in a crisis are snapshots. Snapshots go stale. Weekly cadence keeps the plan accurate.

Treating utilization as the whole picture. Utilization is a measurement, not a forecast. Running at 85% utilization is a fact about last quarter. What utilization will look like in eight weeks, and whether you can live with that number, is a more useful question.

Confusing the resourcing meeting with the planning conversation. Resourcing meetings handle next week. Planning conversations handle next quarter. They pull in different people for a different purpose. When you merge them, the urgent always wins.

If you recognize three of those, you are running a perfectly normal agency.

If your weekly resourcing meeting hasn’t really changed in three months, the conversation that’s actually missing is the one upstream of it. We’ve sat in plenty of those meetings. We’d be glad to walk through what the forecast-first version of yours could look like.

A resource plan is a hypothesis about the next ninety days: a way to know quickly when it stops holding.