A plan to double revenue implicitly assumes the business can deliver double the volume. In many firms that assumption fails before the demand does.
The constraint is usually one thing
Capacity is set by the tightest step in the chain rather than by the average. A single machine, room, licence or qualified person defines the maximum output.
Adding capacity elsewhere produces nothing. The extra work accumulates in front of the constraint, appearing as backlog rather than as revenue.
Identifying which step is actually binding is therefore the first useful act. Investment made without that knowledge tends to buy capacity that was never short.
Capacity arrives in lumps
Physical capacity is rarely divisible. A second production line, a larger unit or an additional shift comes in a fixed size that does not match demand precisely.
Adding one therefore creates a period of expensive underuse followed by a period of comfortable operation, and then shortage again as demand catches up.
This stepped pattern is why capacity decisions are made against forecasts rather than orders. The commitment precedes the demand that justifies it.
Human capacity expands slowly
Where the constraint is skilled people, it cannot be bought at short notice. Recruitment takes months and competence takes longer still.
New staff also consume capacity while learning, drawing on the time of the experienced people who are already the constraint. Output can fall before it rises.
Businesses that grow faster than they can train tend to see quality decline, which generates rework and consumes yet more of the scarce capacity.
Utilisation and responsiveness trade against each other
Running the constraint close to full capacity looks efficient and makes the operation fragile. Any disruption has no spare capacity to recover into.
Lead times lengthen sharply as utilisation approaches the limit, and customers experience the business as unreliable well before it is technically full.
Deliberately holding some headroom costs measurable money and buys unmeasured resilience, which is why it is usually the first thing removed under cost pressure.
Demand can be shaped instead
Where capacity cannot expand quickly, the alternative is changing the demand pattern. Pricing by time of day, lead-time premiums and scheduled delivery all move load away from the peak.
Shifting a modest share of demand off the peak can raise usable output substantially, because the constraint is defined by the peak rather than by the total.
This is generally far cheaper than building for a peak that occupies a small part of the year, though it requires customers to accept a less immediate service.