One-Way Doors: Recognizing Irreversible Decisions

Jeff Bezos made a distinction that seems simple but carries profound temporal implications.

Some decisions are reversible—you can walk back through the door if you don't like what's on the other side. Other decisions are irreversible—once you've walked through, the door locks behind you. You can't go back.

Bezos called them Type 1 and Type 2 decisions. Type 2 decisions (reversible) should be made quickly by individuals or small groups. Type 1 decisions (irreversible) deserve more deliberation, more input, more caution.

The distinction is fundamentally temporal. It's about the arrow of time—about which choices close off futures and which leave them open.

The Physics of Irreversibility

Irreversibility isn't just a decision-making framework. It's a feature of time itself.

In physics, irreversibility emerges from entropy—the tendency of systems to move from ordered to disordered states. You can scramble an egg, but you can't unscramble it. You can shatter a glass, but the pieces won't reassemble. The arrow of time points in one direction.

Some organizational decisions share this quality. Once made, they change conditions in ways that can't be undone:

• Layoffs. You can rehire, but you can't unhappen the layoff. Trust, relationships, institutional knowledge—gone permanently.
• Public statements. Words spoken can be apologized for but not unsaid. The statement exists forever.
• Mergers. Companies can be separated, but not cleanly. The integration creates entanglements that don't fully unwind.
• Product launches. You can discontinue a product, but you can't erase that customers experienced it, that competitors responded to it, that the market's memory includes it.
• Reputation damage. Trust destroyed doesn't rebuild to its original state. Something is permanently different.

The irreversibility isn't always obvious in the moment. The door looks like it swings both ways—until you try to go back.

The Reversibility Illusion

Organizations often underestimate irreversibility. Several cognitive patterns contribute:

The undo assumption. Digital interfaces have trained us to expect reversibility. Control-Z. Restore from backup. Rollback the deployment. But organizational reality isn't a text editor. Most actions leave traces that can't be deleted.

The apology fallacy. "We can always apologize later." But apologies don't restore prior states. They're new events that acknowledge irreversible change. The relationship post-apology is not the relationship pre-offense.

The optionality illusion. "We're keeping our options open." But some choices close options permanently. The decision not to enter a market in 2010 can't be revisited in 2010. That moment is gone.

The iteration assumption. Agile methodology emphasizes iteration—try, learn, adjust. This works for reversible decisions. But truly irreversible decisions don't iterate. You get one pass.

What We're Observing

When we look at organizations through the lens of irreversibility, certain patterns become visible:

The speed trap. Fast decision-making is generally good—but not for one-way doors. Organizations that optimize for speed uniformly may move quickly through doors they can't reopen. The question isn't "how fast can we decide?" but "what kind of decision is this?"

The pilot limitation. "Let's pilot it" works for reversible initiatives. But some things can't be piloted—their effects are immediate and permanent. A pilot layoff isn't a pilot; it's a layoff.

The precedent accumulation. Organizational precedents are often irreversible. Once you've made an exception, the exception exists. Once you've allowed a behavior, you've established that it's allowable. You can change the rule going forward, but you can't undo the precedent.

The relationship asymmetry. Building relationships takes time; destroying them can be instant. This asymmetry is temporal: construction is slow, destruction is fast, and destruction is often irreversible.

The knowledge loss. When people leave organizations—especially when they leave badly—knowledge goes with them. You can hire new people, but you can't recover what the departed knew. Institutional memory is irreversibly diminished.

The Temporal Decision Framework

Recognizing irreversibility changes how decisions should be made:

For reversible decisions (two-way doors):

• Decide quickly
• Delegate to individuals or small groups
• Accept imperfection; you can adjust
• Err on the side of action
• Treat analysis paralysis as the primary risk

For irreversible decisions (one-way doors):

• Slow down
• Seek more input
• Consider second-order effects
• Acknowledge what's being foreclosed
• Treat premature action as the primary risk

The challenge: correctly identifying which type of decision you're facing. The door doesn't come labeled.

Signs of a One-Way Door

We're observing some patterns that suggest irreversibility:

Relationship destruction. If the decision damages trust, relationships, or reputation, it's likely irreversible. These things don't restore to prior states.

Public commitment. If the decision will be visible externally—to customers, markets, media—reversing it carries its own costs. The original decision plus the reversal creates a different state than never having decided.

Legal or contractual binding. Some decisions lock in legally. Contracts, regulatory filings, public offerings—these create obligations that don't simply unwind.

Physical instantiation. Decisions that manifest in physical reality (buildings built, products shipped, infrastructure deployed) are harder to reverse than decisions that remain conceptual.

Knowledge departure. Any decision that causes people to leave—especially experienced people—involves irreversible knowledge loss.

Market timing. Some decisions are irreversible because the moment is irreversible. The decision not to act in a particular market window can't be revisited when the window has closed.

What to Look For

In your own context, you might observe:

• Whether decisions are explicitly categorized by reversibility
• How much deliberation varies between one-way and two-way doors
• Where the organization has walked through doors it now wishes it could reopen
• Whether speed is uniformly valued or calibrated to decision type
• What irreversible decisions are currently being treated as reversible
• Whether "we can always change it later" is being said about things that can't actually be changed later

The question irreversibility raises isn't "is this a good decision?" It's "can we undo this if we're wrong?"

If the answer is no, you're at a one-way door. Slow down.

Temporacy is investigating the hidden temporal structures that shape organizational life. Irreversibility offers one lens for seeing what's usually invisible: which decisions close off futures permanently. We're exploring what this means for organizations navigating choices they can't unmake.