How I Decide When to Slow Down

Speed is not a choice most teams make consciously. It’s what happens when no one pushes back.

The default in most work environments is motion: respond quickly, ship fast, decide and move. That rhythm feels like competence. And sometimes it is. But I’ve learned to distrust the feeling — because it’s indistinguishable from the sensation of burning through credit you don’t know you have.

The question I keep returning to is simpler than it sounds: how do I know when to slow down?

Not in the abstract. Specifically. What’s the signal?

The thing I notice in my own work

There’s a small sensation I’ve started paying attention to — a kind of internal lag before I commit to something. A beat where I almost continue and then don’t.

I’ve started treating that beat as information.

It’s not anxiety. It’s not perfectionism. It’s closer to a sensory pause — the kind that happens when your hand is about to sign something and some part of you is still reading. When I ignore it, I usually find out later that it was pointing at something real: an assumption I hadn’t tested, a step that looked obvious but wasn’t, a consequence I’d glossed over.

When I honor it, I usually find… nothing. A false alarm. Two minutes of re-reading and then I move on.

That asymmetry matters. The cost of pausing unnecessarily is small. The cost of not pausing when you should have is often irreversible — a published error, a misread relationship, a decision that forecloses options you didn’t realize you needed.

The false binary

Most conversations about speed frame it as a tradeoff: go fast or be careful. Ship now or get it right.

That framing is lazy. What it actually describes is two different error modes, and the real work is identifying which error you can afford.

Fast-and-wrong is tolerable when: the decision is reversible, the stakes are contained, you’ll get clear feedback quickly, and the cleanup cost is yours to absorb.

Slow-and-right matters when: you can’t take it back, someone else absorbs the cost of your mistake, or the feedback loop is long enough that you won’t know you were wrong for months.

Most people don’t make this distinction explicitly. They have a default — either toward speed or toward care — and they apply it everywhere. The skill isn’t having the right default. It’s knowing when your default is wrong for the situation.

What I’ve learned to watch for

I don’t have a formula, but I have a short list of conditions that reliably trigger my slow-down reflex:

When I’m moving fast specifically to avoid feeling uncertain. Speed and avoidance feel identical in the moment. If I notice I’m hurrying to get past a decision rather than through it, that’s a flag.

When the answer feels obvious but I can’t explain why. Intuition is often right. But intuition that can’t survive a thirty-second verbal walkthrough is usually not intuition — it’s pattern-matching on a surface feature that doesn’t apply here.

When I’m about to do something that affects someone who isn’t in the room. The people most impacted by a fast decision are often the ones furthest from where it gets made. That gap is where most of the damage accumulates.

When I’ve been moving fast for more than two hours straight. There’s a degradation curve in decision quality that most people don’t account for. I don’t know exactly where mine is, but I know it exists, and I know sustained velocity without a checkpoint is how I end up fixing things I shouldn’t have broken.

The prediction I’ll stand behind

By end of 2026, I think most teams that operate at any meaningful scale with AI-assisted workflows will have explicit “deceleration triggers” built into their operating playbooks — specific conditions that require a pause, a check, a second read. Not because they’re slow, but because they’ve learned the hard way that some mistakes are very cheap to prevent and very expensive to fix.

Teams that don’t build this in will look fast for a while. Then they’ll have a bad quarter and spend three months asking why.

What I still don’t know

I have no idea how to transfer this to other people’s contexts. The internal-lag signal I described is mine — I’ve trained it over years of noticing when I was wrong and tracing it back. I can describe it, but I can’t hand it over.

What I suspect is that the signals are different for everyone and that most people have them already. They just haven’t named them. The work isn’t developing new instincts — it’s learning to trust the ones you already have, which first requires admitting that ignoring them has cost you something.

That’s harder than it sounds.