Skip to content

Marginal Capacity Value Review

Recurring review ritual — instantiates Internal Capacity Deepening

A recurring review that names the currently binding constraint, prices the marginal value of relieving it, and re-ranks relief priorities as the bottleneck moves.

Capacity investment goes wrong when it keeps relieving last quarter's bottleneck. Marginal Capacity Value Review is the standing ritual that prevents this: on a set cadence it identifies which constraint is binding now, prices what one more unit of relief there is actually worth — the constraint's shadow value — and re-ranks where the next increment of effort or money should go. Its defining idea is that this value is marginal and range-bound: relieving the binding constraint is worth a lot right up until the constraint migrates elsewhere, at which point further relief there is worth almost nothing. So the review is inherently recurring — the answer expires as the bottleneck moves, and the ritual exists to re-solve it before stale priorities drive spend.

Example

A hospital's operating-room suite reviews its binding constraint monthly. This month it is not OR time at all — it is recovery (PACU) beds: finished surgeries stack up because patients can't move out of the theatre, so the theatres idle waiting. The review prices a marginal PACU-bed-hour by the throughput it unlocks, and marks the range over which that value holds — a handful of added bed-hours, after which the constraint jumps to anaesthesia staffing and the next PACU bed is worth little. Relief options are ranked by that shadow value against their cost, and a modest PACU-staffing change is prioritized over the capital OR expansion someone had been championing. A quarter later the review finds the constraint has migrated to sterile processing, and the priorities are re-solved accordingly.

How it works

  • Find today's binding constraint first. Every cycle re-identifies which resource actually limits throughput now, before any pricing — because the honest answer changes.
  • Price the shadow value over its range. The marginal value of relieving the constraint is estimated together with the range over which it holds, so no one over-invests past the point where the constraint moves.
  • Re-rank by value per cost. Relief options are ordered by shadow value against cost, so the next increment goes to what binds, not to what was championed.

Tuning parameters

  • Cadence — monthly, quarterly, event-triggered. Match it to how fast the constraint drifts: too slow and priorities go stale, too frequent and it becomes ceremony.
  • Shadow-value method — a formal dual value from a capacity model, or structured judgement. More rigor buys defensible numbers at the cost of effort and false precision.
  • Range width — how far the marginal value is assumed to hold before a re-solve is required. Narrow ranges are honest but demand more frequent review.
  • Scope — the single binding constraint, or the top few. Single keeps focus (the classic discipline); top-few hedges against a fast-moving bottleneck.
  • Standing quorum — who must be in the room for the re-ranking to have authority to redirect spend.

When it helps, and when it misleads

Its strength is keeping capacity investment aimed at the constraint that actually binds now, and stopping over-investment once relief has moved the bottleneck — the marginal value having fallen to near zero the moment the constraint migrates.

Its failure modes come from mis-reading the constraint. Teams elevate a non-binding constraint because it is visible or politically salient, or keep chasing yesterday's bottleneck out of momentum, or treat a shadow value as if it held far outside its valid range. The remedy is the review's own first step, done honestly: re-identify the binding constraint every cycle before pricing anything, and subordinate other decisions to it rather than optimizing everywhere at once.[1]

How it implements the components

  • marginal_response_range — prices the shadow value of relieving the binding constraint and marks the range over which that value holds before the constraint migrates.
  • opportunity_cost_review — ranks relief options by value-per-cost, making the next unit of relief compete against its best alternative use.
  • review_cadence — sets the recurring rhythm so priorities are re-solved as fast as the binding constraint actually drifts.

It does not sense the constraints live or map where they sit (overload_and_underutilization_monitoring, constraint_migration_monitor) — it consumes those from the Network Capacity Dashboard; nor appraise discrete funded investments end to end (capacity_yield_model, fixed_cost_map) — that's Capacity Investment Analysis.

References

[1] The Theory of Constraints (Goldratt) — throughput is governed by the single binding constraint, so the leverage is to identify it, exploit and subordinate to it, then elevate it, and repeat as it moves. The recurring "as it moves" is exactly why this review is a ritual and not a one-off study. ↩