Reversible vs. One - Way Decisions

Picture two doors in a hallway. One is a heavy steel door with a sign “No re - entry.” If you go through it, you can’t come back the way you came; you’d have to deal with whatever’s on the other side or find a new exit. This is a one - way door. The other is a swinging saloon door, clearly marked “push, pull - come and go as you please.” That’s a two - way door; if you step through and don’t like what you see, you can turn around and walk right back out. Now translate this to decisions in work and life. Some decisions are like the steel door: hard or impossible to reverse once made (example: selling your company, firing an employee, launching a product and charging customers money - reputationally hard to undo). Others are like the swinging door: easy to change your mind (example: setting a meeting agenda, trying out a new software tool on a trial, hiring a contractor where if it doesn’t work out, you part ways easily). Recognizing which type you’re facing is crucial, because it dictates how you should approach the decision. Two - way (reversible) decisions can be made quickly, with less analysis, because the cost of a mistake is low (you just backtrack). One - way (hard to reverse) decisions merit more caution, more input, perhaps a pilot test or contingency planning, because you’ll live with the consequences for a long time. Unfortunately, organizations and people often mix these up: they agonize for ages on small reversible things (“bike - shedding” trivial choices), yet sometimes rush into big irreversible commitments without due diligence. Let’s correct that by consciously classifying decisions and tailoring our decision process accordingly.

Classify each decision at hand as two - way or one - way. When a decision comes up, ask: If we make this choice and it turns out wrong, how easily can we undo it or mitigate the damage? If the answer is “we can reverse course pretty simply or at low cost,” then it’s a two - way door. If the answer is “reversing would be painful, expensive, or impossible,” it’s a one - way door. For example, choosing a brand logo for a new product might feel big, but in early stages it’s reversible (you can rebrand before huge public presence). However, deciding to acquire another company is likely one - way (once legal and teams merge, can’t just roll it back without major fallout). Some decisions lie in between; use judgement - if it’s more one - way than not, treat it as one - way (err on side of caution). It’s not always obvious, so discuss with your team: “How reversible is this?” This conversation itself is useful. You might discover, “Oh actually, if customers hate the new feature, we can roll back to previous version in a day - it’s reversible.” Or, “If we move our data to this new platform and it fails, moving back would take months - that’s a one - way move.” Once classified, tag the decision clearly in your documentation or thinking: Write “Type: Two - Way Door (reversible)” at the top of the meeting notes, or explicitly say “This is a one - way decision, folks, so we’ll take more time on it.” This sets expectations on how thorough to be. One - way doors should see more rigor, sign - off from higher - ups, or more evidence gathered, because you want to be quite sure (or have contingency plans if it goes bad). Two - way doors can be delegated down and decided quickly because the risk of a misstep is contained. Amazon’s Jeff Bezos famously used this concept to speed up innovation: many decisions are two - way doors, so make them fast. Only a few are one - way doors requiring slow deliberation. If you treat every decision as one - way (high stakes), you’ll be paralyzed; if you treat one - way as casually as a two - way, you court disaster at times. So the classification step is critical and should become habit.

Set different processes for two - way vs one - way. For two - way door decisions, adopt a bias for action and a lightweight process. That might mean a brief discussion or just an email from the owner: “I’m deciding X, if anyone has objections speak now, otherwise I proceed.” It might mean doing a quick experiment rather than deep analysis. For instance, if you can test a change with a pilot group and easily revert, go ahead and test rather than over - debate it in meetings. Also, empower individuals or small teams to make reversible decisions without requiring consensus from everyone. Speed is the advantage here - since you can fix mistakes, it’s more costly to wait than to course - correct. You might even timebox the decision: “Let’s try solution A for one week. If it doesn’t improve metrics, we swap to B.” Contrast that with one - way decisions: implement a more robust process. Require a written proposal, gather data (maybe base rates, expected value as we learned), and do a pre - mortem (imagine failure and how to mitigate). Often for one - way, it’s wise to involve more stakeholders or higher authority in decision because the stakes are high. Not to spread blame, but to ensure perspectives are covered and buy - in is there, since you can’t revert easily. Set a longer timeline: give people 24 hours or a few days to review the pre - read, insist on an explicit green light from the decider after weighing input. Some organizations have formal “Decision Records” for big one - way moves, documenting why and who decided, helpful because you might need to remember those reasons if years later you’re stuck with the outcome. Also for one - way, design kill criteria and review checkpoints even if you can’t fully reverse. For example, launching a new product line might be one - way (brand and resources commit), but you can set a checkpoint like “If after 6 months sales < 50% of target, we stop further investment.” That’s not fully reversing, but it stops deeper losses. Or in hiring a key position (which if wrong is painful to replace), set a 3 - month evaluation and a plan if it’s not working out (e.g., have a probation period or a potential consultant overlap to pick up slack). Essentially, because you can’t easily undo, you embed caution and exit strategies upfront. Another key for one - way decisions: ensure communication of the rationale and plan widely, since people will have to live with it. You want alignment and understanding to support the irreversible move.

Design cheap tests for reversible choices. When you recognize a decision is reversible, ask, “Is there a quick, inexpensive way to test this before fully committing?” Usually yes. Because it’s reversible, tests themselves are possible without major consequence: that’s the door being two - way. For instance, if considering a new policy in your team, implement it as a “trial run” for a week. If it fails, revert next week - no big harm. If trying a new feature, do an A/B test with a small user percentage; if the feature bombs with them, most users never saw it (reversed quietly). The idea is to take advantage of reversibility to gather real - world data, not just analysis. This complements the EV and probability thinking: probabilities get refined after a test. Say you are considering entering a new market - that might seem big one - way (once brand is out, etc.), but you can often construct a reversible mini - test, like soft launching in one city or with a subset of products. That changes a portion of the decision to two - way: you can exit that city if it flops relatively quietly. For any two - way door, plan a small - scale or short - duration experiment if feasible. It limits downside and provides evidence. And because such decisions should be made quickly, incorporate testing into that speed: instead of meeting for hours, spend those hours on a rapid prototype or pilot. Testing doesn’t always mean actual external actions; it could be simulation or roleplay. E.g., reversible decision to shift team roles - test by swapping roles on a small project for a week, see how people feel. The motto: “If it’s easy to change back, then by all means change and see what happens!” That yields learning and momentum. Make sure though that everyone knows it’s a test and reversible, so they don’t resist as if it’s permanent. You can reduce anxiety by explicitly calling reversible decisions “experiments” or “trial period.” People are more willing to go along since they know if it’s terrible, it’s not forever.

For one - way doors, prepare and document thoroughly. When you decide on a one - way door, treat it like a significant event. Write down why you believe this path is worth the irreversibility. It could be a short memo: “We are doing X because [benefit], we considered alternatives A and B, but chose X due to [reasons]. We acknowledge risks Y and Z; mitigations are [plan]. We believe this bet is worth it because [expected value or strategic need].” Also, define what success looks like and what failure looks like down the road. This mirrors earlier chapters (problem definition, base rates, pre - mortem). By documenting it, if later people question or memories fade, you can revisit the logic and see if assumptions changed. It also helps psychologically to commit: once you’ve stepped through, you go all - in executing since turning back isn’t an option. However, you can pre - commit to an evaluation point: e.g. “We cannot fully reverse, but we can decide at year - end to scale down or pivot strategy if metrics show the decision isn’t delivering as expected.” Have those governance points to avoid throwing good money after bad just due to sunk cost. And assign someone (or a group) to the role of “monitor and adjust.” That is, part of their job is to keep an eye on how this irreversibility is playing out and voice if course corrections around it are needed. You might label them as the “door guard” - empowered to say “hey, maybe this door wasn’t so great, here’s how we adjust forward since we can’t go back.” Often in irreversible decisions, people double down blindly because no one wants to raise hand and say “this isn’t working” since they feel can’t undo it anyway - a phenomenon called “escalation of commitment.” By explicitly assigning someone to look critically, you at least pivot strategy or stop loss beyond a certain point. Many companies have done damage by not having clear kill criteria for projects that turned out to be dead ends due to irreversibility (“We already invested $10M, can’t stop now” thinking). Instead, upfront say “if metric X stays below Y for 3 quarters, we will exit that business line,” which is a one - way door partial reversal (closing it behind you and moving on a different way). Accountability and clarity help because one - way decisions can fail too; the difference is you need to have a plan since you can’t simply revert.

Avoid over - analyzing reversible decisions. A common waste is spending equal time on trivial decisions as big ones because processes are not differentiated. You see teams scheduling large meetings, gathering heaps of data to decide on, say, what color to paint an office wall or which task management tool to trial - things that could easily be changed later. Recognize when you’re stuck in analysis paralysis on a two - way door. Check yourself: “What’s the worst that happens if we make a quick call here and it’s wrong? Can we change it? Yes? Then let’s decide now, go.” Institute time limits for reversible choices. For example, give the team 48 hours to pick a new software tool to test, with the rule that “if it doesn’t work, we’ll try a different one next month.” That speeds up decision by removing fear of finality. Culturally, celebrate quick pivots on two - way decisions as a good thing (“Glad we tried and learned fast!”) to encourage taking those shot. Similarly, avoid rushing or closing debate too soon on one - way decisions in the name of agility. You might hear someone using “bias for action” wrongly on a one - way item which could lead to irreversible mistakes. Push back: “This one, if we get it wrong, we can’t easily undo. It’s worth taking an extra week to gather data.” It’s about matching tempo to door type. Some companies formalize this: minor (two - way) decisions can be made in one meeting by mid - level manager; major (one - way) decisions require a strategy review and sign - off by execs. That’s one way - but even in personal decisions, have that mental sign: “One - way? Sleep on it, consult a friend perhaps. Two - way? Trust my gut and try it today.” This framing reduces decision fatigue because you can quickly filter where to invest your thinking energy. Save deep thought for those one - way doors.

Empower quick reversals when needed. For reversible decisions, not only decide them fast, but also make it safe and okay to reverse them fast if needed. That means ego needs to be checked: if an employee tries a new process (two - way) and two days in says “This isn’t working, I want to go back,” encourage that honesty and swift pivot. Don’t punish, like “Well you chose it, now live with it” - that would effectively turn a two - way door into a one - way by culture, which is counterproductive. Instead say, “Great, we learned that path was suboptimal, let’s revert and try something else.” The quicker you can revert a reversible decision at signs of trouble, the less cost incurred. Perhaps set a policy: any team member can yell “stop” on a trial within the first week if they see clear negative effects, and the team will revert without blame. Or designate a “reversal owner” who is charged with the mechanics of undoing if needed (like a rollback plan in tech deployment). Knowing there’s a plan to reverse easily also encourages people to take bold reversible steps - they have a safety net. Another tactic: simulate the reversal before fully implementing the decision. For instance, if deploying new software company - wide (which you consider reversible if it fails), test the uninstall or data export process as part of trial. That ensures you truly can back out with your data intact. It’s like practicing unlocking the door behind you in case you need to run back. This often uncovers that a so - called reversible decision might have hidden one - way aspects (e.g., a database migration might technically allow rolling back, but it’s painful - so maybe treat as one - way and thus test more or run parallel systems until sure). That nuance is key: sometimes things appear reversible but aren’t fully (due to costs or secondary effects), so adjust classification accordingly.

Label your top five decisions by door type. A quick practice: write down five decisions you or your team are currently mulling or recently made. Mark each as (R) if largely reversible or (O) if largely one - way. For each R, ask “have we perhaps overcomplicated this? Could we just decide now or run a quick test?” If yes, act - schedule that test, or make the call and move on. For each O, ask “have we taken appropriate steps? Is the right person making this decision? Did we stress - test our thinking, plan for downsides, get needed buy - in?” If not, what’s one thing you should do before finalizing it? Maybe you set up a review meeting with a mentor or boss for that hire decision, or do a pilot somehow, or gather one more piece of critical data. Also, check if any decision’s door type is misclassified. If you marked something R (reversible) but on second thought the fallout would be harder than assumed, treat it more cautiously. Or if one feels O (one - way) but you realize a partial retreat is possible, maybe you can de - risk it more. Then, for one of your current big decisions (O), articulate to relevant people “This is a one - way door, so we’re going to be careful - I will make the decision next month after doing X and Y.” And for a small one (R), set an example by deciding quickly: “We’ve discussed enough; this is reversible, so let’s just try Option 2 for now. If it doesn’t work this week, we’ll switch.” This demonstrates the principle in action and often people find it refreshing how it frees them or adds appropriate seriousness. Keep repeating this classification habit until it becomes second - nature. Soon you’ll notice you automatically relax on small stuff (“just do it, we can change it back”), and automatically sit up and think harder on big stuff (“this one counts, let’s double - check our strategy”). You’ll be both faster and safer in decision - making - a combo that’s hard to beat.

You’ve now equipped yourself with a sense of which decisions to sweat and which to sprint through. By aligning decision effort with decision reversibility, you save time and reduce risk across the board. This framework pairs well with all the previous tools: define problems clearly so you know what you’re deciding, use base rates and EV for the heavy one - way calls, use probabilities and checklists to hash things out, then decide swiftly or deliberately as appropriate. With that, we close the section on “Decision Science You Can Use.” Your decisions should now be clearer, better informed, and more appropriately handled for their nature. Next, we’ll turn to principles from physics that you can feel in your daily work - concepts like leverage, inertia, and flow - to further optimize how you execute on those decisions and get things done effectively.