Operations & Supply ChainPlanning & Forecasting

AI Prompts for Capacity Planning

Capacity planning has a translation problem at its core: demand arrives in units and revenue, capacity exists in hours, shifts, machine cycles and square meters. The work is converting one into the other honestly — using demonstrated rather than nameplate capacity, accounting for mix, changeovers and yield — and then deciding how far ahead of demand to add capacity and by which means.

These prompts handle the conversion, the gap options and the roadmap. They rely entirely on the capacity data you provide; a model has no idea what your line actually runs at, and it should say so rather than assume a utilization figure.

Before you use these

Have these ready to replace the highlighted [variables]:

Demand plan by product or family and period
Routing or processing time per unit at each resource, including changeover time
Demonstrated capacity per resource (actual output, not theoretical), shift pattern, planned downtime
Lead times and costs of each capacity lever

The prompts

1. Calculate capacity requirements from the demand plan
2. Size and cost the gap-closure options
3. Build a multi-year capacity roadmap

1. Calculate capacity requirements from the demand plan

Best forA clear resource-by-period load versus available capacity.

Inputs needed

Demand plan
Processing and changeover times
Demonstrated capacity and shift pattern

How to use itGive demonstrated capacity, not nameplate. If you only have nameplate, tell the model and ask it to apply an OEE assumption you supply.

Expected outputLoad vs capacity table per resource and period, utilization, and the first period each resource is overloaded.

You are a capacity planner converting a demand plan into resource requirements.

Demand plan: [product/family × period, units]
Routings: [product → resource, run time per unit, changeover time per batch, typical batch size, yield %]
Resources: [resource, demonstrated capacity (hours or units per period), shift pattern, planned maintenance]

1. Compute required hours per resource per period: (units ÷ yield) × run time + (batches × changeover). Show the formula applied for one product as a worked line.
2. Compare to available hours. Report utilization per resource per period.
3. Flag periods above [threshold, e.g. 85%] utilization as at-risk and above 100% as infeasible. Identify the first infeasible period per resource and the cumulative shortfall in hours.
4. Identify the binding constraint (the resource that limits total output) and whether it moves between periods as mix changes.
5. Sensitivity: how much demand growth or mix shift the binding resource can absorb before it breaks.

State every assumption about yield, changeover and downtime. If demonstrated capacity was not provided and you used nameplate, say so prominently — nameplate overstates real capacity.

2. Size and cost the gap-closure options

Best forComparing overtime, extra shifts, outsourcing, productivity improvement and capital investment on the same basis.

Inputs needed

Capacity gap by resource and period
Constraints and costs of each lever

How to use itInclude lead times. The right option for a gap in month 3 is different from the right option for a gap in month 18.

Expected outputOption comparison with cost per unit of capacity, lead time, flexibility and risk, plus a recommendation by gap horizon.

Act as an operations manager evaluating how to close a capacity gap.

Gap: [resource, periods, shortfall in hours or units]
Levers and their constraints: [overtime: max hours/week, premium %; additional shift: hiring lead time, training time, fixed cost; outsourcing/subcontract: available capacity, unit cost, quality risk; productivity/OEE improvement: achievable %, time to realize; capital equipment: cost, lead time, capacity added]

For each lever:
- Capacity added per period and when it becomes available versus when the gap binds
- Cost per unit of capacity (one-off and recurring separately)
- Flexibility: how quickly it can be reversed if demand falls
- Risks: quality, labor, supplier, execution
- Interaction with other levers (e.g. overtime fatigue reducing OEE)

Then:
1. Recommend a combination for the short-term gap (inside the lead time of structural options) and for the sustained gap.
2. State the break-even utilization or volume at which the capital option beats the variable options.
3. Identify the decision that must be made now because of lead time, and the ones that can wait for better demand information.

Do not recommend a lever that exceeds a stated constraint. Where a cost is missing, give the range that would change the recommendation.

3. Build a multi-year capacity roadmap

Best forPlanning capacity additions against uncertain long-range demand with an explicit timing strategy.

Inputs needed

Long-range demand scenarios
Capacity increment sizes and lead times
Cost of excess capacity vs cost of shortage

How to use itGive the model demand ranges, not a single line. The choice between lead, lag and match strategies depends on that range.

Expected outputRoadmap with timing of each addition, the strategy applied and why, and decision points tied to demand milestones.

You are building a capacity roadmap for [site/network] over [horizon].

Demand scenarios: [low/base/high by year]
Current capacity: [units/year, demonstrated]
Expansion increments available: [option, capacity added, capex, lead time, minimum efficient scale]
Economics: [contribution margin per unit lost when short; carrying cost of idle capacity; capex hurdle]

1. Choose and justify a timing strategy per phase — lead (build ahead of demand), lag (build after demand proven), or match (increments track demand) — based on the ratio of shortage cost to idle-capacity cost and the demand uncertainty.
2. Lay out the roadmap: year, expected utilization under each demand scenario, expansion triggered, decision date (start date minus lead time).
3. Define the demand milestone that must be reached by each decision date to proceed, and the fallback if it is not.
4. Show cumulative capex and the year each increment breaks even under the base scenario.
5. Identify the point of no return for each decision and what information would be worth waiting for.

Present the roadmap as a table. Flag any period where even the low scenario exceeds capacity — that is a commitment, not a decision.

Related prompts

Logical next step

After this, most operations teams move on to Demand Forecasting.

Get the free Operations & Supply Chain AI Starter Kit → Nine of these prompts as a diagnose → analyze → plan workflow with an intake worksheet, delivered by email. See what's inside

All Operations & Supply Chain prompts · Search the full library