What it means
Capacity is the maximum rate at which an organisation can produce output: units per hour, patients per day, transactions per second, billable hours per month. Demand is what customers want, and capacity planning is the discipline of keeping the two matched.
That is harder than it sounds, because capacity is usually added in lumps (a machine, a shift, a building) while demand moves continuously, and because both are uncertain. The process starts with a demand forecast over the planning horizon, expressed in the same units as capacity and including seasonality and growth.
It then measures current effective capacity, which is less than theoretical capacity because of maintenance, changeovers, absence, quality losses and the practical impossibility of running at 100% without queues. The gap between forecast demand and effective capacity, period by period, is the planning problem.
Three broad strategies exist for closing it. A lead strategy adds capacity ahead of demand, accepting under-utilisation for a time in exchange for never turning customers away, and suits businesses where lost sales are costly.
A lag strategy adds capacity only when demand has proved itself, accepting lost sales and overtime in exchange for lower risk, and suits businesses with expensive, specialised capacity and uncertain demand, while a match strategy adds capacity in small increments to track demand closely. Alternatives to adding capacity are often cheaper.
Effective capacity can be raised by reducing changeover time, improving reliability, removing bottlenecks, extending hours or adding shifts. Demand can be shaped by pricing (off-peak discounts), by lead times, or by turning away low-margin work, and peaks can be outsourced or handled with temporary staff.
The financial evaluation uses the same tools as any investment: the cost of the capacity, the contribution from the sales it enables, the cost of the sales lost without it, and the risk that the demand does not appear. The asymmetry matters, because the cost of excess capacity is visible and continuing (depreciation, interest, rent, idle staff) while the cost of a shortage is often invisible (the order that went elsewhere, the customer who did not come back).
Good capacity planning makes both costs explicit. For service businesses and professional firms, capacity is people and the planning is headcount: how many staff, with which skills, are needed to deliver the expected work at a target utilisation, given hiring lead times and attrition.
For IT, capacity is servers, bandwidth and storage, and the planning increasingly uses cloud services to convert lumpy capacity into a continuous, variable cost.
In practice
Real-world examples
Example 1
A hospital plans bed capacity against projected admissions, using average length of stay and a target occupancy of 85% to allow for surges.
Example 2
A software company plans server capacity for a product launch, provisioning cloud capacity to three times expected load for the first week and scaling down after.
Example 3
An accountancy firm plans headcount for tax season, combining permanent staff at 80% utilisation with seasonal contractors for the peak.
Think of it
Capacity planning is figuring out how much you can produce and whether that's enough for demand.
Formula
Calculation
Capacity Requirement = Forecast demand / Target utilisation
Capacity Gap = Capacity requirement minus Effective capacity available
Effective Capacity = Theoretical capacity x Availability x Performance x Quality (the elements of overall equipment effectiveness)
Worked example. A bottling plant has one line with a theoretical capacity of 12,000 bottles an hour, running 16 hours a day, 250 days a year: 48,000,000 bottles a year. Effective capacity, after availability of 88%, performance of 92% and quality of 98%, is 48,000,000 x 0.88 x 0.92 x 0.98 = 38,084,000 bottles (rounded). Demand is forecast at 34,000,000 this year, growing 12% a year: 38,080,000 next year, 42,650,000 in year three, 47,770,000 in year four, 53,500,000 in year five.
Target utilisation is 90% of effective capacity to allow for peaks, so usable capacity is 34,275,000. The plant is at its limit now and will be short by about 3,800,000 bottles next year, 8,400,000 in year three and 13,500,000 in year four. Contribution per bottle is $0.25, so the lost sales without action are 3,800,000 x $0.25 = $950,000 in year two, 8,400,000 x $0.25 = $2,100,000 in year three and 13,500,000 x $0.25 = $3,375,000 in year four.
Options:
- Option A, a second line: capital $9,000,000; adds 38,000,000 of effective capacity; annual fixed cost (depreciation, staff, maintenance) $2,400,000. It needs $2,400,000 / $0.25 = 9,600,000 extra bottles a year just to cover its fixed cost.
- Option B, a third shift on the existing line: adds 8 hours a day, raising effective capacity by 50%; incremental cost $1,500,000 a year including a 15% night premium on that shift's labour; no capital.
- Option C, improve effectiveness: a $1,200,000 programme to raise availability to 94% and performance to 96% lifts effective capacity to 42,449,000 (usable 38,204,000), which covers year two demand of 38,080,000 in full.
Option C removes the whole year-two shortage, worth $950,000, and cuts the year-three shortage from 8,400,000 to 4,450,000 bottles, worth a further 3,950,000 x $0.25 = about $988,000. It repays its $1,200,000 cost by the middle of year three.
After Option C, the remaining shortage in year three of 4,450,000 bottles is worth $1,112,500, less than the $1,500,000 cost of a full third shift, so weekend overtime is the better answer for that year. In year four the shortage is 47,770,000 - 38,204,000 = 9,566,000 bottles, worth about $2,390,000, which justifies the third shift at $1,500,000. Usable capacity with both C and B is about 57,300,000, enough until year six.
The plan: Option C now, overtime in year three, Option B from year four, and a decision on Option A in year five, when demand of 53,500,000 approaches the usable limit and the case for $9,000,000 of capital will be clearer. The lag strategy is chosen because the capital is large, the demand growth is uncertain, and cheaper options buy time.Case study
Seen in the real world
This is a fictional, illustrative scenario. A manufacturer of specialist packaging had turned away $4,000,000 of orders in a year because its main press was full, and the sales director pressed for a second press at $6,500,000. The finance director asked for a capacity study first. It found that the existing press ran at 61% overall equipment effectiveness: changeovers between jobs took an average of 95 minutes and happened eleven times a day, and unplanned stoppages consumed 8% of running time. A six-month programme of changeover reduction (standardised tooling, pre-staged materials, a two-person changeover crew) cut the average to 35 minutes, and a preventive maintenance schedule halved stoppages.
Effectiveness rose to 78%, which added 28% to effective capacity, enough to absorb all the turned-away orders and the following year's growth. The programme cost $400,000. The second press was deferred for three years, at which point demand had grown enough to justify it at 70% utilisation from the first year rather than 45%. The finance director's paper recorded that the company had been about to spend $6,500,000 to buy capacity it already owned, and the board adopted a rule that no capacity investment would be considered until effectiveness of the existing asset had been measured and improved.
Watch out
Common mistakes
- Planning against theoretical capacity rather than effective capacity, which overstates what the operation can deliver by 20% to 40%.
- Adding capacity before improving the effectiveness of what exists, which buys capacity the company already owns.
- Counting only the visible cost of excess capacity and ignoring the invisible cost of shortage, or the reverse.
Questions
People also ask
What utilisation should we plan for?
Below 100%: typically 80% to 90% for manufacturing, lower for services with variable demand, because queues and delays rise sharply as utilisation approaches the limit.
How far ahead should we plan?
As far as the lead time to add capacity plus the time to see demand coming. A factory needing two years to build requires a three- to five-year plan; cloud capacity can be planned in weeks.
What is the difference between capacity planning and capacity utilisation?
Planning decides how much capacity to have. Utilisation measures how much of it is being used.
Watch