Speed vs. Soundness: A Designed Dilemma, Not a Fate

Every manager knows the trade-off, or believes they do. You can decide quickly or you can decide well, and the faster the environment moves, the more you are forced to sacrifice soundness for speed. It presents itself as a law of nature, a tax that reality levies on anyone who must act under pressure. But much of what looks like an inherent trade-off is in fact an artifact—a consequence not of the world's structure but of the organization's, and specifically of what the organization has failed to connect.

Start with why speed usually does cost soundness. A decision made in haste is unsound when it cannot draw on what the organization already knows, cannot be checked against evidence quickly enough to matter, and cannot be corrected before its consequences harden. Each of these is a failure of connection. The knowledge exists but is not retrievable in time; the means of measurement exist but are not wired to the decision; the capacity to adjust exists but is not engaged until far too late. Speed exposes these disconnections mercilessly, and we misread the exposure as proof that speed and soundness are fundamentally opposed.

Boyd's work on decision tempo, later compressed into the idea of the OODA loop, pointed at the resolution without fully naming it. Boyd's insight was that the advantage goes not to whoever is fastest in the abstract but to whoever can cycle through observation, orientation, decision, and action more effectively—learning and re-deciding faster than the situation changes. The point is subtle and usually lost: the goal is not raw speed but a fast loop, one in which acting and learning are tightly coupled so that a quick decision is also a quickly correctable one. Speed and soundness stop being opposites when the decision is embedded in a cycle that catches and fixes its errors rapidly.

This reframes the dilemma entirely. If a decision is treated as a final commitment, then speed genuinely does trade against soundness, because a fast final commitment is a fast bet you cannot take back. But if a decision is made as a provisional move whose results are measured quickly and fed back into the next move, speed becomes compatible with soundness, even a contributor to it—because acting sooner means learning sooner, and learning sooner means correcting sooner. The fast organization is not the one that decides recklessly; it is the one whose decisions are wired to rapid feedback.

What blocks most organizations from this resolution is that the necessary pieces sit in separate compartments. The people who decide are not tightly coupled to the people who measure; the measurement, when it arrives, is slow and oriented toward justification rather than learning; the memory that could have informed the decision is not retrievable at decision speed; and the machinery for adjusting course engages only at the next planning cycle, quarters later. Each compartment may function well internally. The dilemma lives in the gaps between them, and the gaps are precisely what no single compartment can close.

The compartmental structure persists because each compartment is locally rational. It makes sense for measurement to be rigorous and therefore slow; for memory to be comprehensive and therefore hard to query at speed; for adjustment to be deliberate and therefore tied to the planning calendar. Every function, optimizing for its own virtue, produces exactly the latency that starves fast decisions of what they need. No one designed the trade-off; it emerged from the sum of sensible local choices, which is why it feels like a law of nature rather than a consequence of arrangement. And it is why exhorting people to “decide faster and better” changes nothing: the obstacle is not their effort but the shape of the structure they work inside.

So the trade-off between speed and soundness is real, but it is largely designed rather than fated. It is the predictable result of an organization whose decision-making, measurement, memory, and adjustment are managed as separate functions rather than connected into a loop that can move quickly because it corrects quickly. Attack any one compartment alone—faster decisions, better measurement, richer memory—and the dilemma persists, because the problem was never in the compartments. It was in the absence of the connections between them. Dissolving the dilemma is not a matter of choosing speed or soundness more wisely. It is a matter of building the connective structure that would let an organization have both—structure that, once again, no individual improvement supplies on its own.