One-Way Doors: Why "Move Fast" Stops Being Good Advice More Often Than You Think

"Move fast" has been unquestioned executive doctrine for so long that most leaders never examine when it stops being good advice. It stops being good advice more often than the doctrine admits. · Juan Vegarra

"Move fast" has been unquestioned executive doctrine for so long that most leaders never stop to examine when it stops being good advice. In technology adoption right now, it has stopped being good advice more often than most executives realize, and the cost of that miscalibration is showing up in the same failure statistics running through the rest of this series: pilots that never scale, transformations that miss their objectives, AI spend that produces no measurable return.

The speed instinct comes from a real and legitimate fear. Competitors moving faster capture the advantage, and being second with a mature, well-integrated version of a capability is sometimes genuinely worse than being first with a rough one. That dynamic is real in some categories. But it gets applied indiscriminately, as a default cultural posture rather than a situational judgment, and the situations where speed is actually the losing strategy are considerably more common than the doctrine tends to admit in an all-hands meeting.

A useful framework, borrowed and adapted

Amazon's well-known internal framework for this draws a distinction between what it calls one-way doors and two-way doors: decisions that are genuinely irreversible, or extremely costly to reverse, versus decisions that can be undone quickly and cheaply if they turn out to be wrong.^[1] Most day-to-day choices are two-way doors and should be made quickly, with a bias toward action, precisely because the cost of being wrong is low and easily corrected. A smaller number of choices are one-way doors — decisions that lock in an architecture, a vendor relationship, a data model, or a public commitment in a way that's genuinely difficult and expensive to unwind — and those deserve real deliberation before walking through, regardless of how much organizational pressure exists to move at two-way-door speed.

The problem in most organizations isn't that this distinction is wrong. It's that almost nobody applies it consistently. Recent research on strategic decision-making describes a common organizational bottleneck: a feature test that could be safely rolled back in two weeks gets routed through the same six-month approval process as an irreversible platform architecture decision, which ends up bottlenecking a large majority of decisions that should move quickly, while occasionally letting a genuinely irreversible decision slip through on the fast track because nobody paused to classify it correctly.^[1] The same research finds that organizations which apply meaningfully different processes to these two categories of decision move roughly five times faster on the large majority of choices that are actually reversible, precisely because they've stopped forcing every decision through the same heavyweight process.^[1]

Where speed becomes a genuine trap

Speed becomes a trap in three specific, recognizable situations, each of which shows up repeatedly across the rest of this series in different clothing. The first is when the underlying capability or vendor category is still consolidating, with a dozen or more credible players and no clear technical or market leader. Moving fast here doesn't capture advantage. It locks the organization into a one-way-door choice that a slower competitor gets to make eighteen months later with considerably better information, at a lower price, with a more mature product, precisely because that competitor correctly classified the decision as a one-way door and treated it accordingly.

The second is when organizational readiness lags meaningfully behind technical readiness — the exact pattern behind most of the digital transformation failure statistics discussed elsewhere in this series. The technology can be deployed quickly, so it is, and the organizational change genuinely required to use it well — new processes, new skills, new incentive structures — gets left to catch up afterward, which it structurally cannot do at the same speed technology deploys at. Moving the technology fast while the organization moves slow doesn't average out to a reasonable pace. It produces a gap that shows up later as failed adoption, and closing that gap after the fact costs considerably more than a genuinely deliberate pace would have up front.

The third, and the one executives are least willing to name honestly, is when speed is being used to avoid a harder conversation — most often about whether a given capability is actually core to competitive advantage or commodity infrastructure, the distinction covered earlier in this series. Moving fast on a build-versus-buy decision that hasn't actually been thought through isn't decisiveness. It's outsourcing a genuinely strategic question to momentum, and calling the outsourcing "bias for action" in the retrospective.

The honest counter-case

None of this is an argument for deliberation as a universal virtue, and organizations that swing too far in that direction create their own well-documented failure mode: the six-month approval process applied to a decision that was never actually a one-way door in the first place, burning the organization's speed advantage on caution nobody needed. The framework only works if the classification step happens honestly and early — is this decision genuinely hard to reverse, or does it just feel important because a senior executive cares about it — rather than defaulting every decision above a certain dollar threshold into the slow lane regardless of its actual reversibility. Misclassifying a two-way door as a one-way door is exactly as expensive as the reverse mistake, just less visible, because the cost shows up as competitors quietly outpacing you rather than as a single dramatic write-off.

The discipline that actually matters

The executives worth betting on aren't the fastest movers or the most deliberate ones. They're the ones who can articulate, for any given initiative, why the current pace is the correct one — because the underlying decision genuinely is reversible and the organization should move quickly, or because it genuinely isn't and deserves real deliberation before committing. Speed is a tool, not a virtue in itself, and the organizations still treating it as a default cultural posture rather than a situational judgment are going to keep showing up in next year's failure statistics, just as quickly as they showed up in this year's. For the biggest technology decision on your roadmap right now — is it actually a one-way door, or does it just feel like one because everyone in the room is nervous?

Juan Vegarra is the author of An Outsider's Playbook (forthcoming). The views here are his own. More essays · Write me