Demand, Capacity & Scenario Planning
Matching supply to demand under uncertainty: forecasts, utilisation, bottlenecks, service levels, inventory policy, how the network responds to a shock, and scenario planning. Forecast error becomes either shortage or write-off, and in regulated supply the two are not symmetrical: the shortage side carries patient harm. Inventory is the buffer that absorbs variability, and it is also the first thing optimised away.
What an explainer is not
A topic explainer is SPEQ’s synthesis of what a practice involves, cited to the standards that govern it. It does not reproduce their text, and it does not determine which of them apply to your product or process.
[ POSITION IN THE FRAMEWORK ]
7 DIMENSIONS · 22 LINKSForecast error is asymmetric in regulated supply: an over-forecast becomes write-off, an under-forecast becomes shortage, and only one of those two reaches a patient.
06 · QUALITY MATURITY — DEMAND, CAPACITY & SCENARIO PLANNING, REACTIVE TO ADAPTIVE
Demand is forecast and production is planned against it. Inventory policy is a target set in months of cover.
Forecast accuracy is measured and inventory is optimised against it, treating over and under supply as symmetric errors.
The asymmetry drives policy: cover is set higher where a shortage would reach patients, and the reasoning is recorded.
Regulated constraints are modelled rather than approximated — qualification lead time, shelf life, release cycles, batch sizes — so plans are executable.
Scenarios rather than a single forecast drive capacity decisions, and the quality organisation is planned against the same demand as production.
SPEQ’s shared five-stage progression, labelled synthesis — not the FDA QMM rating scale. Where does your organization sit? Score your quality system →
07 · REGULATORY & EVIDENCE
GOVERNING STANDARDS · 4
Derived from the 4 standards SPEQ maps to this subject, across 3 regulatory bodies: FDA, ICH, ISO.
RECORDS & OBJECTIVE EVIDENCE
- Inventory policy with the reasoning for cover levels, including patient-impact weighting
- Constraints modelled in planning: shelf life, release cycles, qualification lead times, batch size
- Scenario analyses supporting capacity decisions
- Forecast accuracy measurement, and what it changed
- Quality and laboratory capacity planned against the same demand as production
COMMON INSPECTION FINDINGS
- Inventory policy set uniformly in months of cover regardless of patient impact
- Plans that assume release cycles or qualification lead times shorter than reality
- Capacity decisions taken on a single forecast with no scenario analysis
- Production capacity increased without corresponding quality and laboratory capacity
- Shelf life ignored in planning, producing write-off that was structurally inevitable
The asymmetry that should drive inventory policy
Standard inventory optimisation balances holding cost against stockout cost, and in most industries a stockout costs a sale. In regulated supply a stockout of a product with no therapeutic alternative costs something the model cannot price, and it may be a reportable event carrying regulatory consequence.
That asymmetry argues for holding more inventory on clinically critical products than a symmetric model would recommend, and it argues for the classification to be clinical rather than commercial. Where inventory policy is set by value, the products protected are the profitable ones — which is a defensible commercial answer and the wrong regulated one.
Regulated supply has constraints ordinary planning does not model
Shelf life caps how far ahead inventory can be built. Batch sizes are fixed by the validated process, so supply comes in quanta rather than continuously. Release testing adds lead time between manufacture and availability that varies with laboratory load. Campaign scheduling means a missed slot may wait months. And a quality event can quarantine finished goods that the planning system counts as available.
A planning model that treats manufacturing as a continuous, immediately available capacity will systematically understate the response time to a demand change. The most valuable single correction is usually to model release lead time as a variable rather than a constant, because it is the term that lengthens precisely when supply is under pressure.
Scenarios beat forecast accuracy
Effort spent making a point forecast more accurate has limited return, because the events that cause supply failures are not forecast errors — they are discrete shocks: a site interruption, a supplier failure, a demand surge from a competitor’s withdrawal, a regulatory action. None of these is captured by narrowing a confidence interval.
Scenario planning asks what the network does under each, and its output is not a prediction but a set of prepared responses and an understanding of which products have none. That second part is the valuable one: knowing in advance which products would be in shortage within weeks of a specific failure is what makes the mitigation decision fundable while there is still time to act.
SPEQ interpretation — plan the quality organisation against the same demand
Demand and capacity planning covers manufacturing capacity, and typically stops there. But a demand increase also increases testing volume, batch record review, deviation load and release decisions, and those capacities are planned separately if at all — which is why a successful demand ramp so often produces a laboratory backlog and a lengthening release cycle.
Extending the capacity model to the quality organisation is not complicated: tests per batch, review hours per batch, and a realistic deviation rate give a load that scales with the plan. Doing it turns a predictable constraint into a planned one, and it is the same omission the capital case makes at a different point in the lifecycle.
FREQUENTLY ASKED
Why should regulated inventory policy be asymmetric?
Because a stockout of a product with no therapeutic alternative costs something an optimisation model cannot price, and may be a reportable event. That argues for holding more on clinically critical products than a symmetric model recommends — and for classifying criticality clinically rather than by value, since value-driven policy protects the profitable products.
What do ordinary planning models miss in regulated supply?
Shelf life capping forward build, fixed validated batch sizes making supply arrive in quanta, release testing lead time that varies with laboratory load, campaign slots that wait months, and quality events quarantining stock the system counts as available. Modelling release lead time as variable rather than constant is usually the highest-value correction.
Why is scenario planning more useful than forecast accuracy?
Because supply failures come from discrete shocks — a site interruption, a supplier failure, a demand surge, a regulatory action — not from forecast error, and narrowing a confidence interval captures none of them. The valuable output is knowing which products would be in shortage within weeks of a specific failure, while mitigation is still fundable.
What capacity is usually left out of demand planning?
The quality organisation. A demand increase raises testing volume, batch record review, deviation load and release decisions, and those are planned separately if at all — which is why a successful ramp so reliably produces laboratory backlog and a lengthening release cycle. Tests and review hours per batch make it modellable.