Early on, I treated every decision like it might be my last. I would gather one more data point, ask one more person, sleep on it one more night. It felt responsible. It felt like the careful, mature way to run a company. What I was actually doing was paying an enormous and invisible tax on speed, in exchange for a sense of safety that was mostly imaginary. Most of those decisions, it turned out, were easy to undo. The thing I could not get back was the time I spent agonizing over them.
The single most useful shift in how I lead came from a simple reframe: decisions are usually cheaper than the delay required to feel certain about them. Most leaders overweight the cost of being wrong and underweight the cost of being slow. Flip that, and a team moves at a completely different speed, not because it is reckless, but because it has stopped paying for certainty it was never going to get anyway.
Most decisions are doors you can walk back through
The cleanest framing I know of separates decisions into two kinds. There are doors that lock behind you, and doors you can walk back through. Selling the company, taking on a huge debt, firing a key person without cause: those lock. Choosing which feature to build first, picking a vendor you can switch, naming a product, structuring a team this quarter: those open both ways. You can reverse them, usually at modest cost, once you learn something.
The error that slows companies down is treating reversible decisions with the same caution as irreversible ones. People bring the gravity of a one-way door to a problem that is plainly a two-way door, and they spend weeks deliberating over something they could have tried, learned from, and corrected in days. Good decision making starts with honestly classifying the door in front of you. The overwhelming majority of choices you face in a week are reversible, and reversible decisions deserve speed, not ceremony.
Once I started naming this out loud in meetings, the tone changed. "This is a two-way door, let us just pick and move" became a normal sentence. It gave people permission to act, which is mostly what they were waiting for.
The cost of delay is real and almost never counted
Here is what I underweighted for years. When you delay a decision to gather more information, you are not choosing between a wrong answer and a right one. You are choosing to operate without an answer for the entire length of the delay. The team waits. Adjacent work stalls because it depends on the call. Momentum leaks. Two other decisions pile up behind this one. None of that shows up on a balance sheet, so we pretend it is free. It is not free. It is one of the largest costs a company carries, and it is paid in silence.
I now try to put a number on delay, even a rough one. If a decision is blocking four people for a week, that week is not contemplation time, it is roughly four person-weeks of stalled work plus the harder-to-measure drag on energy and pace. Against that, the cost of picking the slightly worse option on a reversible call is often trivial. When you actually compare the two, the case for deciding now becomes obvious in a way it never is when delay feels like the cautious, free default.
Example: Picture a team spending eleven days choosing between two analytics vendors that are 90 percent identical. The "wrong" choice would cost maybe a few hours to switch later. Those eleven days of meetings and comparison cost far more than switching ever would, and the team could have learned which was better by simply using one.
Speed compounds, and so does hesitation
The reason this matters beyond any single call is that the habit compounds. A team that decides quickly on reversible things makes far more decisions per quarter, which means it learns far faster, which means its later decisions are better informed. Speed is not just about moving faster today. It buys you more cycles of learning, and learning is the thing that actually improves judgment over time.
Hesitation compounds in the other direction. A team that deliberates over everything not only moves slowly, it teaches itself that deciding is dangerous. People start routing choices upward to avoid the risk of being the one who chose. Decisions queue at the few people allowed to make them, and the whole company starts moving at the speed of its most cautious meeting. I have seen genuinely talented organizations grind nearly to a halt this way, not from any single bad call, but from the accumulated weight of calls nobody would make.
How to actually decide faster without being reckless
Moving fast is not the same as moving blind. The point is to spend your caution where it belongs and stop spending it where it does not. A few rules have served me well, and they are most of what I would tell a newer leader about prioritization and pace.
- Classify the door first. Reversible or not. If reversible, set a tight deadline and decide by it. If irreversible, slow down on purpose, because that is where slowness earns its keep.
- Decide with 70 percent of the information. Waiting for 90 usually costs more than the extra 20 is worth. The last bit of certainty is the most expensive and the least decisive.
- Name the owner. Every decision needs one person who decides. Shared decisions are how things sit for weeks while everyone waits for someone else.
- Set a default. If no decision is made by the deadline, what happens automatically. A good default turns indecision into a choice instead of a stall.
We bake this into how we run the company and into how Atlas is built. Clear ownership on work, deadlines that actually mean something, and a written record so that when a reversible decision needs reversing, you can see what you knew at the time and change course without drama. The record matters because it removes the fear: if deciding fast is safe to undo, people decide fast.
The mindset shift, in one line
If I could install one belief in every leader I work with, it would be this: you are almost never as trapped by a decision as you feel, and you are almost always paying more for the delay than you would for the mistake. The careful instinct that tells you to wait is not wrong about everything. It is just badly calibrated, applying one-way-door caution to two-way-door problems. Recalibrate it, decide quickly on the reversible majority, save your deliberation for the rare locks, and your whole team will feel the difference inside a month. Decisions really are cheaper than you think. It is the waiting that quietly drains the account.