The Goal Question

In the AI Story

The Goal Question's power is diagnostic. Most management decisions that produce disappointing results do so because they optimize something other than the goal — a local metric, a departmental objective, an efficiency measure that feels important but does not correspond to system purpose. Goldratt's insistence on the question is a discipline against this drift. Before evaluating any improvement, ask: does this move the system toward the goal? The answer, honestly given, often reveals that the proposed improvement is locally rational and systemically useless.

The generalization beyond manufacturing is: produce value now and in the future. Value is the thing someone outside the system is willing to exchange something for. Not output. Not activity. Not the appearance of productivity. Value. The measurement framework follows: throughput (rate of value generation), inventory (investment in convertible assets), operating expense (cost of maintaining conversion capacity). Every decision that fails the test — that increases throughput at the cost of proportionally greater inventory, or decreases operating expense at the cost of greater throughput loss — moves the system away from the goal.

Applied to the AI transition, the Goal Question produces a dramatically more nuanced picture than the dominant narratives allow. Operating expense has decreased — unambiguously. Potential throughput has increased — also unambiguously. But potential is not actual. Actual throughput is determined by the constraint, which is now judgment. And inventory — cognitive, digital, product — has accumulated at unprecedented rates as the AI's generative capacity outpaces evaluative capacity. The goal after the orange pill is the same goal as before: produce value now and in the future. But the definition of value has ascended along with the constraint. When coordination was the constraint, value was partially filtered by implementability — the coordination budget determined what could be built. Now, with coordination gone, the filter must be performed by judgment: should this be built?, not just can this be built?

Segal's formulation in The Orange Pill — 'Are you worth amplifying?' — is, in the final analysis, a Goal Question. It asks whether the builder's judgment, amplified by AI, will produce throughput or inventory. Value or waste. Products that serve genuine needs or products that serve the builder's ego, the market's short-term appetites, or organizational inertia that mistakes activity for progress. The constraint has always been judgment. For fifty years, the coordination bottleneck hid this fact. The AI broke the alibi. Judgment stands exposed — not because AI changed what judgment is but because AI removed the infrastructure that concealed what judgment was not.

Origin

Goldratt introduced the Goal Question as the organizing inquiry of The Goal (1984). Alex Rogo's journey toward answering it — rejecting 'satisfy customers,' 'improve quality,' 'reduce costs' in favor of 'make money now and in the future' — is the narrative spine of the novel. The question has since become a fixture of TOC-informed management practice and has been extended to non-profit and public-sector contexts with appropriate modifications to the definition of value.

Key Ideas

What is the goal? The question sounds banal until attempted with precision, at which point it becomes the hardest question in management.

Most standard answers are not goals. 'Reduce costs,' 'improve quality,' 'satisfy customers' are necessary conditions or partial indicators, not goals.

Profit is oxygen, not purpose. The stark formulation — make money now and in the future — names the necessary condition without which nothing else can be sustained.

Value generalizes beyond profit. Produce value now and in the future — value being what someone outside the system will exchange for.

The AI transition requires returning to the Goal Question. When the constraint shifts, the definition of value ascends with it: should this be built? replaces can this be built?