The Supply Line in Capacity Planning
The supply line in capacity planning is the stock of capacity that has been decided on but has not yet arrived: requisitions open, offers accepted, classes in training, and graduates still ramping to full productivity, and, where decisions are committed ahead of release, those decisions too. It is the workforce planner's version of the "orders placed but not yet received" that the beer-game experiments showed decision-makers systematically underweight,[1] and ignoring it is a common structural cause of the hiring cycles that operations attribute to forecast error. The page defines the stock, gives the decision rule that counts it, shows what happens when it is left out, and sets out what a planning process needs in order to count it.
The page belongs to the series that begins at Workforce Planning as a Control System.
The stock, defined
A capacity plan closes a gap between the agents it wants and the agents it has. The question is what "has" means. Three readings are in use, usually without anyone noticing which one a given report takes.
| Reading | What is counted | Who uses it |
|---|---|---|
| Agents on the floor | people seated and productive, sometimes weighted for ramp | the operation, intraday, and most headcount reports |
| Agents on the floor plus people on the way | the above plus trainees, accepted offers and open requisitions | the recruiting report, and the hiring decision in a well-run capacity cycle |
| Approved positions | the above plus decisions taken but not yet released, inside a budget envelope | finance, in the position-control view |
The supply line is the difference between the first reading and the second, weighted for the attrition it will suffer before it lands and the productivity it will deliver when it does. In a service workforce where the pipeline is two to four months and annual attrition is in double digits it is not a rounding error: at a replacement rate of half a percent a week and a twelve-week pipeline the steady-state supply line is about six percent of the floor, and during a ramp it can be a quarter [estimated].
The decision rule that counts it
System dynamics gives the stock-management structure as the standard form of the hiring decision.[2] In planning terms:
- hires this period = expected losses + (plan of record − agents on the floor) ÷ adjustment time − w × (supply line − desired supply line) ÷ supply-line adjustment time
The first term replaces the people who will leave. The second closes the gap to the plan, over an adjustment time rather than at once. The third is the supply-line correction: when more is already on the way than the replacement pipeline needs, hire less. The desired supply line is the expected losses multiplied by the pipeline length, which is what has to be in flight at all times just to stand still.
The weight w is the whole argument. At w = 1 the rule counts everything on the way. At w = 0 the rule ignores the supply line entirely. Subjects in the beer-game experiments behaved as though w averaged about a third (0.34 in secondary reports of the fitted decision rules).[1] A capacity plan that computes its gap as "plan minus people on the floor" has set w = 0 whether or not anyone intended to.
The same weight belongs in the cancellation rule. A policy that ignores the pipeline when it hires also ignores it when it decides it has too many, so it cancels requisitions it will need in a quarter; a policy that counts the pipeline nets the replacement flow out before it calls anything excess. Reviewing requisitions on the same cadence as the hiring decision, rather than every week against a monthly batch, is what keeps a monthly cycle from cancelling its own replacement hiring.
What happens without it
With w = 0 and the whole gap hired each month, a step increase in demand produces the following sequence. Month one: the gap is hired. Month two: nothing has arrived, the gap is unchanged, the gap is hired again. Month three: the same. Month four: the first batch arrives, the gap closes, and two more batches are still in the pipeline. Months five and six: the floor overshoots, requisitions are frozen or cancelled, and the surplus is paid for. Attrition during the freeze drains the surplus, the gap reopens, and the sequence repeats. The period of the cycle is set by the pipeline, not the forecast; in the companion simulation it runs at roughly three times the pipeline length for a twelve-week pipeline, and lengthens as the pipeline is lengthened [estimated].
The cycle is visible in data an operation already has. Requisitions opened and cancelled in alternate quarters. Attrition during a freeze described as unexpected. Overtime that peaks a quarter after a hiring decision. A headcount series that, once the seasonal calendar is removed, has a dominant period of several pipelines (about three in the companion simulation [estimated]).
Counting the supply line is necessary but not sufficient. In the simulation, adding the supply-line term to the monthly rule reduces the ringing after a step but does not remove it, and does not by itself lower cost, because the whole gap is still being hired at once; the ringing goes when the adjustment time is lengthened to about eight weeks on a twelve-week pipeline and only variance that has persisted is released. That is consistent with the theory, where stability depends on the supply-line weight and the adjustment time relative to the delay together.[2] A hypothesis the model suggests but does not test is that counting the supply line in a cascade that still anchors its forecast to the annual plan makes an operation stably late rather than oscillating: it stops overshooting toward a number that is itself wrong [estimated].
Counting it in practice
Four things are needed, none of them analytical.
- One supply-line report, by pool. Decisions taken, requisitions open, offers accepted, classes in training by week of graduation, graduates by week of ramp. These are often held in different systems and rarely summed.
- A ramp curve per pool. A trainee is not an agent and a graduate is not a tenured agent; the supply line is counted in productive-equivalents, not heads.
- Attrition applied to the pipeline. Trainees leave, often faster than tenured staff. The supply line that lands is smaller than the one that was opened.
- The gap computed net. The capacity cycle's deficit is plan minus floor minus landing supply line. Where Capacity Planning Cycle describes the supply roll-forward, this is what the roll-forward must contain.
Limitations
The decision rule is a standard form, and real hiring decisions are made by committees on partial information with their own delays; the rule describes the structure, not the meeting. The experimental weight of about a third comes from a laboratory supply chain, not from a workforce. The simulation's periods and amplitudes are illustrative. And counting the supply line does not help if the plan of record it is counted against is wrong, which is why this page is one of a series rather than a remedy on its own.
Maturity Model Position
At the lower levels of the WFM Labs Maturity Model™ the supply line is a recruiting report and the gap is plan minus floor. Level 4: Planning in Distributions counts it in the deficit by pool with ramp and attrition applied, and Recruiting Pipeline and Capacity Planning describes the pipeline it reads.
See Also
- Workforce Planning as a Control System
- Feedback Loops in Workforce Planning
- Recruiting Pipeline and Capacity Planning
- Capacity Planning Cycle
- Planning Nervousness and Frozen Zones
- Systems Thinking
References
- ↑ 1.0 1.1 Sterman, J. D., "Modeling managerial behavior: misperceptions of feedback in a dynamic decision making experiment", Management Science 35(3), 1989, pp. 321–339.
- ↑ 2.0 2.1 Sterman, J. D., Business Dynamics: Systems Thinking and Modeling for a Complex World, McGraw-Hill, 2000, chapter 17, "Supply Chains and the Origin of Oscillations"; the labor supply chain is chapter 19.
