Resource forecasting for agencies: a practical guide

Many agencies currently track everything in email and Excel: a resourcing request lands in an inbox, someone checks a spreadsheet that was last updated two weeks ago, and a guess goes back out. That process works until the pipeline moves faster than the spreadsheet does, which is most of the time. This guide covers how to forecast resource needs for the next quarter without that lag: the cadence to run it on, what to read to separate real signal from noise, and how to move the process off email and spreadsheets.

What resource forecasting means for agencies

Resource forecasting is projecting the people, skills, and hours you will need over the coming weeks or months, based on work that is likely but not yet confirmed. It is different from capacity planning, which checks whether your current team can deliver what is already committed. Forecasting looks further out and deals with less certain inputs: pipeline deals, renewal patterns, and planned initiatives instead of signed statements of work. Our guide to resource forecasting vs capacity planning covers the distinction in more depth if you need the full breakdown.

The output of forecasting is a simple answer: for the next quarter, where will you be short, and in which skill. That answer is only useful if it arrives early enough to act on, which is the part a spreadsheet-and-email process struggles with.

The forecasting cadence: weekly and monthly

Run a light weekly check and a deeper monthly review. The weekly check is a quick read of what changed: which deals moved, which projects extended, which people came free. It should take minutes, not hours, because most weeks nothing dramatic shifts. The monthly review is where you actually re-forecast: walk the pipeline by likelihood, map it against team availability by skill, and update the quarter's picture.

A quarterly-only cadence is too slow. Pipeline moves week to week, and by the time a quarterly review catches a shift, the hiring or redeployment decision it should have triggered is already late. A weekly-only cadence without the deeper monthly pass tends to drift into reacting to whatever changed last week rather than holding a real quarter-out view. Both cadences are needed, and both depend on having current numbers to look at, which is where a spreadsheet-based process usually falls behind: someone has to remember to update it, and during a busy stretch that update is the first thing to slip.

Reading availability against committed work

Forecasting compares two numbers for each person or role: hours already committed, and hours available. Committed hours come from confirmed allocations, plus a weighted share of pipeline deals likely to close. Available hours come from contracted hours minus approved leave and any standing non-billable commitments.

The comparison only works if both numbers are current. A spreadsheet usually gets this wrong in one of two ways. Either it shows committed hours from confirmed work only, which understates real demand because it ignores the deals about to close, or it treats every pipeline deal as certain, which overstates demand and triggers a hiring decision you did not need. The fix is weighting: apply each deal's probability of closing to the hours it would need, so a deal at 30% likelihood contributes 30% of its hours to the forecast rather than all of them or none.

Read the comparison by skill rather than by headcount alone. A firm can look perfectly staffed in aggregate while one discipline, say backend engineering or a specific design skill, is heavily over-committed and another sits idle. A skill-level breakdown surfaces that gap; an aggregate headcount number hides it.

The signals that predict over-allocation

A handful of signals show up before a team is actually over capacity, and catching them early is the point of forecasting at all:

• A single skill or role shows committed hours climbing across two consecutive monthly reviews, even if the aggregate team number still looks fine.
• Several deals with the same required skill sit in the pipeline at similar close dates, which concentrates demand into a narrow window even if the total looks manageable spread across the quarter.
• A person's committed hours (confirmed plus weighted pipeline) approach or exceed their available hours more than four weeks out, before anyone has actually said yes to the work.
• Renewal or retainer work that historically renews at a predictable rate is not yet reflected in the forecast, understating demand you can reasonably expect.

None of these signals require certainty to act on. The point of forecasting is to move a hiring, redeployment, or pipeline decision earlier, while there is still time to act on it, rather than waiting for the over-allocation to become a scheduling emergency.

Moving off the spreadsheet

The forecasting process itself is not complicated. What breaks it in practice is where the inputs live. A spreadsheet-and-email approach depends on manual updates from several sources, and forecasting is only as current as the slowest one.

Each row in that table usually sits in a different tool, which is exactly why email and spreadsheets end up as the connective tissue between them: someone has to manually pull pipeline data, cross-reference leave, and check the roster, then reconcile it all into one view. That reconciliation is where forecasts go stale, because it takes real effort and gets skipped or delayed the moment the team gets busy.

Moving off the spreadsheet means connecting those inputs so the forecast updates as the underlying data changes, rather than waiting on someone to rebuild it. A pipeline deal moving to 70% likelihood should shift the forecast without a manual edit. Approved leave should come out of available hours automatically. This is also where forecasting connects back to capacity planning: our resource capacity planning guide covers the near-term side of the same problem, and the two work best run together rather than as separate exercises with separate spreadsheets.

Where Pike fits

Pike keeps allocations, availability, and pipeline in one workspace, so a forecast reflects committed work and weighted pipeline without a manual reconciliation step. Leave comes out of available hours automatically, and the view breaks demand down by skill, so a gap in one discipline doesn't hide behind a healthy aggregate headcount number. See how it works on the resource management feature page, or see your team's real-time capacity and forecast what's coming: book a 15-minute walkthrough.

Frequently asked questions