Supply and Attrition Forecasting (generic process)

Supply and Attrition Forecasting (generic process) is the generic process card, GP-105, for the recurring process by which a workforce function projects the headcount it will have at each future period: the roster of record rolled forward with attrition by cohort, the hiring pipeline discounted by its ramp, internal moves, and leave and return, published as effective FTE with bands and graded assumptions. Its product is the supply side of the capacity gap; it feeds stage 2 of the monthly cycle every month and step 5 of the annual cycle once a year. The modeling is documented on Workforce Forecasting (cohort flow, survival and ramp models, the practitioner playbook) and on Employee Attrition and Turnover (attrition drivers, predictive analytics, nesting survival curves); neither is restated here. This card carries the operating process, who rolls what forward and when, with the L0 profile, the variant classes, the L3 job aids and the automation candidates. It is part of Process Standardization and sits under the forecasting section of the function catalog in Anatomy of a ROC Standard.

The process is separate from the demand forecast because its mathematics are different: demand is a time series, supply is a stock-and-flow model of people whose departure hazard depends on tenure and whose productivity depends on time since hire. It is run on the demand forecast's cadence and horizon nonetheless, because a supply forecast on a different clock makes the gap stale.

Process profile

The profile extends the L0 identity fields of Process Decomposition (L0–L3) with the process-level facts a standards committee needs before any step is drawn. Local values are held by the deploying function and marked SET LOCALLY.

Field	Generic value
Card	GP-105 · local card id assigned by the standards committee at deployment
Class	strategic, in the catalog's four-class vocabulary (strategic · real-time · both · supporting)
Trigger	the capacity cycle's calendar: the roll-forward opens when the roster of record closes for the month, at a business-day offset (SET LOCALLY) ahead of stage 2; an out-of-cycle run on a trigger (a class cancelled or short, attrition beyond a band, SET LOCALLY)
Cadence	monthly on the capacity cycle's horizon; quarterly refresh of the attrition and ramp curves; annually on the long-range horizon for GP-101
Scope in	the opening balance by cohort, skill, pool and status; attrition by cohort; the hiring pipeline by class; ramp; promotion, transfer and pool moves; leave and return; the net effective FTE by period with bands; the accuracy of last period's supply vintage
Scope out	the demand forecast; the requirement and the gap (stage 3 of GP-102); the hiring decision and the requisition; the retention program that moves the drivers; the reconciliation method itself (Headcount Reconciliation in Service Operations)
Inputs	the roster of record; the headcount of record from finance; separations and leave records from human resources; recruiting's pipeline (class dates, sizes, offer acceptance); training's cohort schedule and graduation actuals; the definitions register (the attrition and shrinkage definitions); the curve library of record
Outputs	the supply vintage (effective FTE by cohort by period, with bands at quantiles SET LOCALLY (P50/P80/P95 is typical [A])); the assumptions sheet with a grade per assumption; the class-graduation projection; the score of last period's vintage with error attributed; the hand-off to stage 3
Upstream	the human-resources system of record; recruiting; training; the placement function for partner seats (committed hours rather than heads)
Downstream	GP-102 stages 2 and 3; GP-101 step 5; scheduling (the pool it may schedule); GP-505 (the skill mix); GP-106 (scoring)
Resources	the roster of record; a cohort-projection model; the curve library (attrition survival per segment, ramp per skill); the reconciliation query. Platforms are named at L3, never here
Decision points	does the roster reconcile to the headcount of record? are the curves inside their refresh window? does the pipeline match the plan? does any assumption carry [A] with no [E] range? has an out-of-cycle trigger fired?
Authority tier	the supply planner builds (tier 1); the planner who owns the client portfolio signs the assumptions sheet and the vintage at the gate (tier 2); a vintage resting on an [A] attrition rate with no range beside it is blocked under Human Gates and Number Grades and reaches the capacity cycle only with a range or a tier-4 (function head) logged acceptance
Risk and compliance flags	individual-level predictors (leave reasons, predicted departures) fall under data-protection law in several jurisdictions, so the forecast works on cohorts and individual risk scores never leave human resources; works councils may have a say in predictive attrition analytics; partner rosters are the partner's under contract
Variant-prone steps	the roster source where no system of record exists; partner seats where supply is hours not heads; leave categories by jurisdiction; the ramp curve where training is partner-run
Automation candidates	eight L2 sub-steps (six ranks), listed below

L1 flow

One owner per step; owners are seats. The forecast works at cohort level throughout; no step consumes an individual's risk score.

1. Close and reconcile the roster: pull the roster of record at the close and reconcile it to the headcount of record. Decision: does it reconcile within tolerance (SET LOCALLY)? If no, hold until the data steward's correction lands (supply planner).
2. Build the opening balance by cohort: tenure band, skill, pool and status (supply planner).
3. Decision: are the attrition survival curves and the ramp curves inside their refresh window? If no, refresh them from the last quarter's exits and the graduation actuals training supplies, before proceeding (supply planner).
4. Forecast attrition by cohort: apply the survival curve per segment, estimated on the censored exit history in the manner of the product-limit estimator, and the labor-market modifier; grade each assumption (supply planner).
5. Load the hiring pipeline: class start dates, sizes and offer acceptance; training's cohort schedule and pipeline attrition are a sub-step (recruiting liaison).
6. Apply ramp: convert each class to effective FTE by period on the ramp curve for its skill, net of pipeline attrition (supply planner).
7. Forecast internal flows: promotions, transfers, pool moves and planned organizational moves, from historical rates and the moves the operations managers have declared (supply planner).
8. Forecast leave and return at cohort level from the human-resources partner's leave records (supply planner).
9. Run the cohort projection: net effective FTE by cohort by period, with bands at quantiles SET LOCALLY (P50/P80/P95 is typical [A]) (supply planner).
10. Decision: does any assumption carry [A] with no [E] range beside it? If yes, obtain a range or send it to the function head for a logged acceptance; if no, proceed (supply planner).
11. Score last period's supply vintage against the actual roster and attribute the error to attrition, pipeline, ramp or flows; record the scores for GP-106 (supply planner).
12. Planner gate: review the assumptions sheet and the vintage and produce a decision per client portfolio (planner).
13. Stamp the vintage and hand it to stage 3 of the monthly cycle with the assumptions sheet; publish the pool to scheduling and the partner-seat view to placement (supply planner).
14. Decision: has an out-of-cycle trigger fired (a class cancelled or short, attrition beyond the band, SET LOCALLY)? If yes, re-run steps 4 to 9 for the affected cohorts and re-issue through the gate (supply planner).

L2 step table

Generic sub-steps only; the deploying function adds the columns Process Decomposition (L0–L3) leaves open. Durations are typical business days [A], SET LOCALLY.

Step	Sub-step	Owner (seat)	Inputs	Outputs	Duration	Automation candidate	Documentation join key
1	1.1 Pull the roster of record at the close	supply planner	the system of record	the roster extract	0.5 BD	Y	roster close query
1	1.2 Reconcile to the headcount of record; hold on a breach of tolerance	supply planner	the extract, finance's count	the reconciled roster	1 BD	Y	reconciliation instruction
2	2.1 Assemble the opening balance by cohort	supply planner	the reconciled roster, the cohort table	the opening balance	0.5 BD	Y	cohort definition table
3	3.1 Test curve currency; refresh from last quarter's exits and graduates if due	supply planner	exits, graduation actuals	the curves of record	2 BD	Y	curve library instruction
4	4.1 Apply the survival curve per segment with the market modifier; grade each	supply planner	the opening balance, the curves	the attrition layer	1 BD	N	attrition assumptions sheet
5	5.1 Confirm class dates, sizes and acceptance	recruiting liaison	the requisition log	the pipeline sheet	1 BD	N	pipeline intake form
5	5.2 Confirm the cohort schedule and pipeline attrition	training lead	the training calendar	the cohort sheet	1 BD	N	pipeline intake form
6	6.1 Convert each class to effective FTE on its ramp curve	supply planner	the pipeline, the ramp curves	the ramp layer	0.5 BD	Y	ramp application instruction
7	7.1 Enter promotions, transfers and pool moves	supply planner	historical rates, planned moves	the flows layer	1 BD	N	flows instruction
8	8.1 Enter leave and return by cohort	supply planner	leave records (cohort level)	the leave layer	0.5 BD	N	leave and return profile table
9	9.1 Run the cohort projection with bands	supply planner	all layers	the supply vintage draft	0.5 BD	Y	projection run instruction
10	10.1 Test every assumption for an [A] with no range; obtain a range or send to the function head	supply planner	the assumptions sheet	the graded sheet	same day	N	assumptions sheet
10	10.2 Accept or refuse an ungraded assumption with a logged reason	function head	the escalated assumption	the logged acceptance	same day	N	assumptions sheet
11	11.1 Score last period's vintage and attribute the error	supply planner	last vintage, the actual roster	the score sheet	1 BD	Y	vintage scoring template
12	12.1 Review the sheet and the vintage; produce a decision per client portfolio	planner	the vintage, the sheet, the scores	the gate record	1 BD	N	planner gate pack
13	13.1 Stamp the vintage	supply planner	the gated vintage	the stamped vintage	15 min	N	hand-off instruction
13	13.2 Assemble the hand-off to stage 3, scheduling and placement	supply planner	the stamped vintage, the sheet	the published vintage	0.5 BD	Y	hand-off instruction
14	14.1 Test the out-of-cycle triggers; re-run the affected cohorts	supply planner	the trigger table, events	a re-issued vintage	SET LOCALLY	N	out-of-cycle runbook

Variants

The known variant classes for this process, recorded per Standard and Variant in Process Documentation:

• Statutory. A jurisdiction that restricts the use of employee-level predictors changes nothing in a cohort-level forecast; the restriction is a parameter on the inputs of step 4, SET LOCALLY. A jurisdiction where a works council must be consulted before predictive attrition analytics are used on its unit takes a variant row at step 4, with the market modifier only and no unit-level predictors, owned by the works-council liaison.
• Contractual. A partner seat whose supply is contracted as hours rather than heads takes a variant row at steps 2 and 9: the opening balance is a committed-hours block and the partner's attrition sits inside its price. Owned by the placement lead, reviewed at contract renewal.
• Structural. A unit with no roster system of record rolls forward from a headcount file; step 1 is varied under a row reviewed each cycle until a system of record exists, as on GP-102. A unit whose training is partner-run with no graduation actuals carries its ramp curve at [A] under a row at step 3.
• Transitional. A unit from a predecessor organization whose human-resources system has not migrated keeps a variant row with expiry at cut-over; its automation effect reads blocks against ranks 1 and 2 for that unit.

L3 register

A localizing team writes these job aids against the generic L2; names only, the content is the function's. The reconciliation tolerance, the curve refresh window, the band quantiles and the out-of-cycle band are L3 values marked SET LOCALLY in the L2 row.

Job aid	Anchored to	Form
The roster close query and the reconciliation instruction	step 1	work instruction; method on Headcount Reconciliation in Service Operations
The cohort definition table (tenure bands, pools, statuses)	step 2	work instruction
The curve library instruction: refresh window, fitting, acceptance	step 3	work instruction; method on Survival Analysis for Workforce Attrition and Speed to proficiency curve
The attrition assumptions sheet with grades	steps 4 and 10	work instruction
The pipeline intake form for recruiting and training	step 5	work instruction
The ramp application instruction	step 6	work instruction
The flows and the leave-and-return profile tables	steps 7 and 8	work instruction
The projection run instruction, with the band quantiles	step 9	work instruction
The vintage scoring template	step 11	work instruction
The planner gate pack	step 12	work instruction
The out-of-cycle trigger path	step 14	runbook

Automation candidates

Flagged at L2 per the method on Process Decomposition (L0–L3); released per class of action on a published catch rate per Human Gates and Number Grades, never on a maturity label. The list is ordered by how easy a catch rate is to publish, and every row marked Y above appears in it.

1. The roster pull and reconciliation (1.1, 1.2): a deterministic query and a sum against finance's count; the same first candidate GP-102 names, because a reconciliation that misses is caught the same day by the count it failed to match.
2. The opening balance (2.1): deterministic given the cohort table; it ranks after reconciliation because a wrong balance is only caught once the roster it is built from has been reconciled.
3. The ramp conversion and the cohort projection (6.1, 9.1): computed given the curves and the pipeline, with the overseer checking the assumptions sheet rather than the arithmetic; they rank after the balance because their errors are only countable when the next period's roster closes.
4. The hand-off assembly (13.2): assembled from the stamped vintage; it ranks after the projection because an error here is caught by the receiving stage's acknowledgement rather than by a rate of its own.
5. Vintage scoring (11.1): computed once the actual roster closes; it ranks after the hand-off because its catch rate needs several periods of scored vintages before it can be stated.
6. The curve refresh (3.1): the fitting is computed, but the acceptance of a refreshed curve stays human until the refresh's own catch rate is published; last because a curve fitted on a small cohort can move the forecast further than the drift it corrects, and that error takes a quarter or more to show.

The planner gate (12.1) and the stamp (13.1) are never candidates: a signature is a human act by definition. Nor is the escalation of an ungraded assumption (10.1, 10.2), because the gate exists so that a person must produce a range or a reason, not acknowledge a flag. The attrition layer (4.1) is deliberately not listed either: the hazard depends on tenure, on the unit and on labor-market conditions the model does not observe, and a century of turnover research has not reduced it to a rule a function could certify.

Maturity Model Position

Four scales on this wiki use the word level; the launch page states which is which. On the WFM Labs Maturity Model™ a function that rolls its roster forward monthly on a reconciled opening balance, with attrition by cohort rather than one annual rate and with ramp as a curve, is at Level 2; one that publishes bands, grades every assumption, scores each vintage and has released ranks 1 to 3 above on a published catch rate is at Level 3; one whose supply vintage is a machine-readable object read directly by the capacity plan and the placement engine is at Level 4.

See Also

• Process Standardization — the hub and the inventory this card belongs to
• Workforce Forecasting — the supply-side modeling this card operates
• Employee Attrition and Turnover — attrition drivers, prediction and the nesting survival curve
• Survival Analysis for Workforce Attrition — the statistical method behind the curves
• Speed to proficiency curve — the ramp curve the pipeline is discounted by
• Headcount Reconciliation in Service Operations — the reconciliation step 1 performs
• Monthly Capacity Planning Cycle (generic process) — the stage this process feeds
• Annual Strategic Workforce Planning Cycle (generic process) — the long-horizon run
• Forecast Accuracy Governance (generic process) — where the vintage scores go
• Standard and Variant in Process Documentation — how the variants above are recorded
• Wiki:Packs/Process Decomposition — the pack that localizes this card (CP-OPS-001)