Capacity Planning and Finance: Turning the Staffing Plan Into the Revenue Forecast

Every services firm has two numbers for the next quarter. Operations says one thing. Finance says another. They reconcile in a spreadsheet nobody enjoys, usually the day before the board meeting. Both sides are working from real data. Both sides are wrong in different ways, because the two numbers were never built to be the same number.

The fix is not a better spreadsheet. It is a reconciled model where the staffing plan is the revenue forecast, and FP&A adds the financial overlay without duplicating the underlying data.

Why do the staffing plan and the forecast disagree?

Three specific structural reasons, at almost every services firm.

• Different granularity. The staffing plan runs at weekly allocation percentages. The forecast runs at monthly revenue totals. Any information the plan captures at week resolution gets averaged away by the time it reaches the forecast.
• Different assumption sets. The plan uses list rates. The forecast uses realized rates after discount. The plan ignores write-offs. The forecast bakes them in. The plan assumes clean invoicing. The forecast assumes lag.
• Different owners. Resource management updates the plan. FP&A updates the forecast. Both are correct within their own view. Neither view is complete.

Nobody is at fault. The two documents were never designed to be one document, so they diverge whenever the underlying reality changes.

What does a reconciled model actually look like?

One data layer, two lenses. That is the structure.

• Data layer. Every allocation for every person for every week, at percentage granularity, tagged with the engagement and the bill rate.
• Operations lens. Resource management sees the grid. Utilization, bench, overallocation, staffing gaps. Updates weekly.
• Finance lens. FP&A sees the same data multiplied out to revenue, layered with write-off assumptions, discount adjustments, and invoicing timing. Updates monthly for the formal forecast, weekly for the internal view.

The critical property is that both lenses derive from the same underlying data. When resource management changes an allocation on Tuesday, the finance view moves too, without a reconciliation step.

How do you handle the assumption gap?

Rates, discounts, and write-offs are the biggest source of divergence between operations and finance. Handle them explicitly.

Operations uses the effective bill rate in the plan. Finance layers write-offs and timing on top. The two views are consistent because the effective bill rate is a shared input, not a re-derivation.

What about non-billable time?

Non-billable time is where a lot of forecasts go wrong. Handle it consistently.

• Structural non-billable time. Training, internal projects, admin. Set at the role level as part of the utilization target. Does not appear in the forecast as revenue, appears in the model as a cost.
• Sales support time. Delivery hours spent on scoping and proposals. Depending on firm policy, either non-billable overhead or partially billable to the client if it exceeds a threshold. Set the policy once.
• PTO. Reduces available hours. Does not appear as revenue. Does appear as a fully loaded cost.

The forecast should show billable revenue in the numerator and cost of delivery, including all non-billable time, on the cost line. Do not net non-billable time out of the revenue side.

How do you catch drift between the two views?

Weekly variance reports. Fifteen minutes each Monday.

• Compare planned revenue for last week against actual. Planned comes from the staffing plan. Actual comes from approved hours times bill rates.
• Variance under 3%. Normal. No action.
• Variance 3 to 7%. Investigate. Usually a scope change or a time tracking error.
• Variance above 7%. Structural issue. Either the plan is optimistic, the time tracking is inaccurate, or the rates have moved without updating the plan.

The point of the variance report is not blame. It is diagnostic. Which of the three inputs (plan, tracking, rates) is currently the least reliable, and what does that suggest about the near-term forecast.

Who owns each part of the reconciliation?

The most important structural decision is clarity on ownership. Four roles, four different responsibilities.

• Resource manager. Owns the staffing plan. Ensures allocations reflect the intent for each week.
• FP&A lead. Owns the forecast model. Ensures assumptions and adjustments are applied consistently.
• Delivery leads. Own scope changes and their reflection in the plan.
• Head of ops or CFO. Owns the joint output. Escalation path when the two views diverge.

If any of these four is unclear, the reconciliation collapses. The most common failure is the CFO expecting FP&A to reconcile alone, without operational input. That is the setup where forecasts stay wrong month after month.

What does the joint operating rhythm look like?

Three cadences, integrated.

• Weekly (Monday, 30 minutes). Resource manager and FP&A lead review last week's actuals vs plan. Identify any variance above threshold. Update the current week's plan.
• Monthly (first workday of month, 60 minutes). Full financial forecast update. FP&A layers write-off adjustments, refreshes discount assumptions, and closes the month.
• Quarterly (last week of quarter, 90 minutes). Base rates review. Write-off percentage recalibration. Role-level utilization targets checked against actuals.

Each cadence produces a specific output. Nothing gets discussed in the wrong meeting. The weekly is tactical, the monthly is financial, the quarterly is strategic.

How does this change hiring decisions?

Hiring decisions should be made against the reconciled forecast, not against either the plan or the forecast alone.

• Signal from the plan. A role is short in the base case for the next 12 weeks. Utilization is above target in that role for the current period.
• Signal from the forecast. The revenue attached to that role is sustained across the forecast horizon, not a one-quarter blip. Write-off adjustments do not eliminate the demand.
• Decision. Hire when both signals agree. Hold when they disagree. Escalate when they disagree and the operational signal is strong.

Most bad services firm hires happen when one signal is strong and the other has not been checked. The reconciled model forces the check.

What actually matters

The staffing plan and the revenue forecast are two views of the same underlying question: what is the firm going to deliver in the next 12 weeks, and what is that worth. When they are treated as separate documents with separate owners and separate cadences, they diverge, and every downstream decision, from hiring to pricing to bench management, gets made against a partial view. The reconciled model does not require new tooling. It requires that both views draw from the same allocation data, that assumptions are shared explicitly, and that variance is treated as a diagnostic signal rather than an accounting problem. The forecast accuracy improvement is the visible payoff. The clarity in decisions is the larger one.