Inventory

In Goldratt’s accounting, inventory is one of three operational measures, alongside Throughput and Operating Expense. Standard ToC defines it as the money the system has invested in things it plans to sell — raw materials waiting to be processed, work-in-progress between steps, and unsold finished goods — and Goldratt later broadened the term to “investment,” covering everything the system buys to enable production. (This triad and the “investment” rename come from Goldratt’s ToC writing and standard ToC references, not from the CF source material, which discusses inventory chiefly through buffers and local optima.) Together the three measures let you judge whether a local change actually helps the global goal of making money, rather than just looking busy.

Crucially, inventory is not an asset to be maximized. Piling up parts looks like productivity but mostly signals misdirected effort. Material released ahead of what the constraint can absorb does not become sellable output faster — it just sits, tying up cash and lengthening lead times. So a rising inventory level away from the bottleneck is a warning sign of local optimization, not progress.

CF draws on this measurement triad as a model of how to reason about any system with a goal, not only factories: the right scorecard prevents mistaking activity for achievement. The remedy described in the CF treatment is to subordinate release to the constraint — keep inventory low except for a deliberate buffer protecting the bottleneck — so resources flow toward the goal instead of accumulating where they cannot yet help. This connects directly to drum-buffer-rope, which paces material release to constraint capacity precisely to stop inventory from ballooning. See also The Goal, Goldratt’s novel where these ideas are introduced.