Headcount scenario builder

A usable headcount scenario names the revenue forecast it is tied to and the cost assumptions that price each role. If either input is missing, those fields stay empty. Do not invent a cost. Do not invent a headcount. FP&A or the HRBP still chooses which scenario, if any, becomes the working plan. A generated draft is not an approval.

That rule exists because a scenario without a vintage is a guess with a table around it. Reviewers cannot tell whether hiring is tied to the published reforecast, an older close, or a file someone renamed. A scenario that fills loaded cost when compensation data never loaded is worse: it looks finished, and finished gets treated as true.

The working method is to load the forecast first, build cited scenarios side by side, leave blanks where inputs are missing, and keep the choice with a human.

Open the planning cycle you intend to staff against. Pull the revenue forecast FP&A published for that cycle, including its vintage: close date, version name, or reforecast label. Pull the cost assumptions that belong to the same cycle: salary bands or midpoints, burden rates, start-month phasing, and any location or level differentials already in the model. If the organization runs a continuous workforce model, use that model's current published snapshot rather than an ad hoc export.

Do not start from a prompt that asks for a hiring path and assume a forecast will attach itself. Natural language is useful for stating the question, for example what a conservative path looks like if you staff only roles already on the approved requisition list, costed against this vintage. It is not a substitute for loading the vintage.

Workforce and planning suites such as Workday, Anaplan, Adaptive, and Oracle HCM hold versions, drivers, and workforce plans. Treat them as a class of systems of record for the numbers you cite. The scenario builder should read the version you name. It should not replace those systems, and it should not be described as if it adds capabilities those platforms do not expose in your tenant.

If the forecast vintage cannot be resolved, stop. Leave cost and headcount empty on every line that depended on that vintage. A blank line is the correct output. Using last cycle's burden rate, or a midpoint someone remembers from recruiting, is inventing a cost.

Attrition is not something you invent inside this scenario. If you need a supply view, pull a cited source such as predictive attrition forecasting. Do not turn an attrition view into a hiring target here. Do not put an attrition rate on the scenario as if it were an outcome of the plan.

With vintage and cost assumptions loaded, generate two or more scenarios from those same inputs so the comparison is honest. Each scenario should show, on the artifact itself:

• Forecast vintage, labeled the way FP&A versions it
• Cost assumption set: which compensation file, which burden version
• Headcount lines only where both a role and a costed rate were available
• Empty cells where a role was requested but cost or forecast coverage was missing
• Phasing that follows start dates already in the plan, not dates the model invented

Side-by-side comparison is the point. One path might hold hiring to roles already on the requisition list. Another might add roles the revenue forecast's growth drivers imply, but only where a costed midpoint already exists in the assumption file. A third might show no hiring in a function whose revenue is not in this vintage. All three must cite the same vintage. If they do not, you are not comparing scenarios. You are comparing different plans that happen to sit in one grid.

Leave blanks. If a requested role has no loaded cost assumption for that level and location, the cost cell stays empty and the headcount for that line stays empty. Do not substitute a national average. Do not pull a figure from live market benchmarking into the scenario unless that file is an official cost assumption for this cycle and is cited as such. A benchmark can inform a later compensation conversation. It is not a silent fill for a missing loaded cost.

Internal supply follows the same rule. If you expect backfill from mobility rather than external hire, cite the internal mobility supply forecaster vintage or leave the backfill line blank. Do not invent a transfer the supply view did not produce.

One walk-through, not a result

An HRBP supporting a product organization meets FP&A after the quarterly reforecast is published. They load that reforecast as the vintage and load the same-cycle compensation and burden file. They ask for three scenarios: hold to open requisitions only; staff roles the growth drivers imply where costed rates exist; show no hiring in a function whose revenue line is not in this vintage.

The builder returns a side-by-side. The requisition-only path lists only reqs that already have a costed rate; reqs with no loaded midpoint stay blank. The growth-driver path adds roles that map to drivers in the reforecast, again only where a costed rate exists, and does not add headcount for a role with no rate. The uncovered-function path shows empty hiring lines for that function instead of a manufactured savings total.

None of the three is the plan yet. The HRBP and FP&A select one, or they send the set back because the cost file was the wrong version. That is the example. There is no implied right headcount and no before-and-after figure.

The quality bar is a scenario that cites vintage and assumptions, with blanks where inputs were missing. A person still chooses. Auto-approval is not part of the workflow.

The usual failure is forwarding the comparison to a staffing meeting as if finance already signed it. The artifact can look official: tables, loaded costs, a vintage stamp. None of that is an approval step. Until a named person selects a scenario and that selection is recorded in the planning system of record, the draft is a discussion object.

A related failure is promoting the path that looks most complete. Completeness often means gaps were filled. Prefer the scenario that shows blanks over the one that looks finished without a vintage.

When you do choose, record the choice: which scenario, which vintage, which cost file, who chose, and the date. Later hiring, offer, and budget conversations should point at that record, not at the prompt that generated the draft.

A scenario with no forecast vintage is not a scenario. If the builder cannot name the revenue forecast version, do not present costed totals. Empty stays empty. Asking the model to use whatever is latest, without resolving which version FP&A published, is how last week's spreadsheet becomes this week's headcount.

Treating the draft as approved creates commitments nobody made. Recruiters open requisitions, hiring managers interview, and finance later says no path was selected. Keep the draft labeled draft until a named person chooses.

Inventing a cost or a headcount to avoid a blank cell poisons the comparison. A missing midpoint, a missing burden rate, or a role with no forecast coverage stays blank. You can retrieve the missing file and regenerate. You cannot estimate a number into the table and still call the scenario cited.

Walk the side-by-side the way an auditor would.

Confirm every scenario header shows the same forecast vintage, or an explicit note that a scenario could not be built because the vintage was missing. Confirm the cost assumption set is named and matches this planning cycle, not a leftover file. Confirm blank cells are still blank. Confirm no attrition rate appears as a hiring outcome. Confirm no total is shown for a column that includes inferred or placeholder costs.

Confirm you still have a comparison: more than one path, same inputs, human choice required. Confirm planning systems were used as sources of versions, not as a reason to trust an uncited export.

This page does not pick the headcount. Whatever gets picked must trace to a forecast vintage and a cost file. Anything that could not be traced stayed empty.