capacity

Capacity is the maximum Throughput your operation can sustain - tickets resolved per day, orders shipped per hour. It is the ceiling on your Value Stream.

The critical insight: the cost to maintain capacity grows faster than capacity itself. Model the relationship as Cost = k × C^α, where C is capacity, k is a cost coefficient specific to your operation, and α is the scaling exponent (α > 1).

Why does cost grow faster than linearly? Three compounding forces:

• •Coordination overhead: Adding people to a team increases communication paths. Five people have 10 paths; fifteen have 105. The formula is C × (C - 1) / 2 - the complete graph, and the theoretical maximum. Real organizations compress well below this through hierarchy, team structure, and tooling. That compression is what management is for. The coordination component still scales faster than linearly, but the exponent depends on how well your organization manages the compression.
• •Error Cost: More people and more volume means more points of failure and higher defect rate. Errors do not just add - they multiply across dependent steps in your Value Stream.
• •Tooling and process: Larger teams need heavier infrastructure - more Quality Systems, more Exception Review, more management layers.

How large is α? It depends on the ratio of coordination cost to capital cost in your Cost Structure. The C-squared model (α = 2) is the theoretical maximum from the complete coordination graph. Most Operations fall below it. Calibrate α from your historical data (see How It Works). When you cannot calibrate, use α = 2 as a conservative planning bound - it overestimates real costs in most Operations, which is the right direction to be wrong.

Minimum capacity: You cannot operate at zero. Even if Demand drops to nearly nothing, you still pay the cost of at least one person, one server, one shift. This floor (C ≥ 1, cost ≥ k) creates Fixed Obligations on your Operating Statement regardless of volume. Entering any line of business commits you to this floor. Setting C = 0 - exiting - is the only way to eliminate it.

Three P&L effects:

1. 1)Cost Per Unit can rise with scale in coordination-heavy Operations. If cost grows as C^α and output grows as C, then Cost Per Unit grows as C^(α - 1). This is not universal. Rough guide: if more than half your Cost Structure is Labor, coordination costs likely dominate and Cost Per Unit rises with scale. If more than half is tooling and production lines, capital leverage likely dominates and Cost Per Unit falls as fixed costs spread across more output.
2. 2)There is an optimal capacity level for any given Revenue. Build too little and you lose Revenue from unmet Demand. Build too much and the superlinear cost eats your Profit. The Operator's job is to find that point.
3. 3)The minimum-capacity floor is a cost you pay regardless of volume. If Revenue cannot cover the floor, the right answer is often exit - not Cost Reduction.

Model: Cost = k × C^α, where C is capacity, k is a cost coefficient, and α is the scaling exponent.

Revenue comes from the Demand you actually serve, which cannot exceed C.

The optimization:

• •You choose C.
• •You pay k × C^α regardless of how much Demand arrives.
• •You capture Revenue on min(Demand, C) - you serve whichever is smaller.
• •Profit = Revenue from served Demand - k × C^α.

Since C ≥ 1, you are always paying at least k. Shutting down a product line (setting C = 0) is sometimes the right call - it is the only way to eliminate that floor cost.

The cost of the next unit of capacity:

Under the C-squared planning bound, going from C to C + 1 costs k × (2C + 1). For other exponents, compute k × (C + 1)^α - k × C^α directly. Either way, the incremental cost grows with your current capacity level. Each additional unit delivers the same Throughput but at escalating cost - diminishing returns on capacity investment.

Decision rule: Add capacity when the expected marginal contribution from one more unit exceeds the incremental cost. Stop when it does not.

In practice, you rarely know Demand perfectly. You make this decision under uncertainty, which means you need Expected Value calculations across Demand scenarios to find the right capacity level.

Calibrating the model to your operation:

If you have total cost data at two known capacity levels (C₁, Cost₁) and (C₂, Cost₂), solve for both parameters:

• •Scaling exponent: α = ln(Cost₂ / Cost₁) / ln(C₂ / C₁)
• •Cost coefficient: k = Cost₁ / C₁^α

If α comes out close to 1, your operation scales nearly linearly - coordination overhead is low relative to direct costs. If α approaches 2, you are near the theoretical maximum. The single most important practical step is knowing your actual scaling exponent from data - not assuming it. The first worked example walks through a full calibration and shows what happens when the model overshoots reality.

Apply capacity thinking whenever you face a scaling decision:

• •Hiring Targets: Adding people to a team. Before approving Hiring Targets, estimate total cost including the coordination overhead new team members create for existing people - not just the per-person cost.

• •Infrastructure: Provisioning servers, warehouse space, production lines. Capital-heavy infrastructure often scales more favorably than Labor-heavy teams, but still has superlinear components in failure modes and oversight requirements.

• •Entering a new market or product line: The minimum capacity constraint means you commit to at least C = 1 cost. Run the break-even analysis: does the Expected Value of Revenue in that market exceed k over your Time Horizon?

• •Deciding whether to exit: If Demand has dropped and you are stuck at minimum capacity, you are paying k for a line of business that may never cover its floor. Compare the ongoing cost to the opportunity cost of redeploying those resources. Sometimes the right capacity is zero - which means exiting.

Capacity decisions are downstream of your Cost Structure analysis - they set the scale at which those costs operate. Downstream, capacity connects to Cost Per Unit (which can rise in coordination-heavy Operations as C scales, since cost grows faster than output), break-even calculations (where the superlinear cost curve shifts your break-even volume nonlinearly), and Bottleneck identification (your operation's capacity is limited by its tightest constraint). The Fixed vs Variable Costs distinction matters directly: variable-cost capacity additions that avoid coordination overhead can capture Demand upside without the superlinear penalty of permanent expansion.