Slow is the new fast.

Slow is smooth. Smooth is fast.

I have carried that sentence for years, and for most of them I understood only half of it. I heard it as permission — a way of dignifying caution, of feeling wise while nothing shipped. It took losing a fair amount of time to learn that the sentence is not about caution at all. It is about sequence.

Because there is a version of this argument that is worthless, and it is the popular one. It says: be patient, good things take time, resist the noise. All true, all useless, because it gives you no way to decide anything on a Tuesday. Every real decision is a question about pace, and "go slower" is not an answer to it. Half of what I do should be done quickly and half of it should not, and getting that division wrong has cost me more than any single bad decision ever has.

So let me try to say the useful version.

First, the contradiction

I should deal with the obvious objection, because it is sitting in the room.

I do not live slowly. I work nearly every day. I sleep four or five hours. If you followed me around for a month you would not describe what you saw as cathedral-building; you would describe it as something closer to a sprint that has forgotten to end.

And I want to be precise about why, because the internet has a whole industry built on selling those hours as a virtue, and that is not what this is.

I keep the hours because I love the process. That is the entire explanation. I resent sleep in the specific way you resent an interruption — it is a gap between me and the thing I want to be doing. I sleep because a body is a constraint and not a suggestion. If it were up to me I would not.

So do not copy the hours. Copy the reason, if you have it. A person who works like this out of love is doing something quite different from a person doing it out of fear, even though the calendar looks identical from outside. And I will say plainly what I said elsewhere: it costs me things — relationships, responsibilities, a certain warmth I used to have. It is a season, not a system.

But here is the point. The builder can run at any pace he likes. The building has its own clock, and it does not care. The vision I am actually working toward takes decades. Nothing I do at two in the morning changes that. What my hours buy me is more attempts inside a long window — not a shorter window.

Confusing those two is the most common mistake in ambitious people. Working furiously feels like acceleration. Usually it is just density.

What speed actually broke

I have paid the tuition on this, so let me be specific about the bill.

Before the business I run now became what it is, we pivoted repeatedly. Not the celebrated kind of pivot that gets written up afterwards as insight — the other kind, where you keep moving because the last thing did not work and you are not entirely sure why. Every one of those pivots traced back to the same root cause. We committed early. We saw an opportunity and moved on it before we understood the whole picture, because moving felt like progress and studying felt like hesitation.

Here is what that gets you. You enter a market, you do the work, and some months in you discover why the market was untapped. It was not waiting for you. It was empty for reasons — structural, economic, unglamorous reasons that were entirely discoverable beforehand by anyone willing to spend three weeks asking uncomfortable questions and sitting with the answers.

That is the thing about due diligence nobody says out loud: the correct outcome of research is often that you do not build. We treat research as a step on the way to launching, so we do it in a way that can only ever produce a green light. But a fortnight of honest questions that stops you entering the wrong market has saved you a year. It will never feel like an achievement. There is no launch, no announcement, nothing to show. You just quietly do not lose the year.

And when you do launch prematurely, the punishment is not usually a dramatic failure. It is scrambling. You spend the following months in permanent reaction — fixing the positioning, re-explaining the offer, patching things that a fortnight of thinking would have got right the first time. Scrambling is dangerous precisely because it feels like effort. You are busy. You are exhausted. You are working every hour. And almost none of it is compounding, because you are paying off a decision you made too early.

Am I arguing for endless planning? No. That is the same disease wearing the opposite coat. Planning past the point of information is just anxiety with a document. Research answers questions; planning past that point invents comfort. The difference is testable: if you cannot say what you are trying to find out, you have stopped researching.

Where I am deliberately fast

Now the other half, which matters more, because this is the part that stops all of this from being a philosophy for people who are simply slow.

When something is still being conceptualised, I move as fast as I possibly can.

Early on, you are not building — you are searching. You do not yet know what the thing is. And in a search, speed is not recklessness, it is the entire method, because the only way to find out what works is to run out of things that do not.

The way I think about it: if you knew for a fact that the yes was a hundred nos away, how would you behave? You would sprint at those hundred nos. You would get excited by each one, because every no is not a rejection, it is a piece of the map, and you have just moved one closer with certainty. Nobody deliberates when the goal is to exhaust a list.

That is how we are building the newer things in our house right now. Move fast, break the idea against reality repeatedly, find out quickly what it actually is. Not because we are impatient, but because the question we are trying to answer — what exactly is this? — cannot be answered by thinking. It can only be answered by contact.

And then, the moment we know: slow down. Deliberately. That is the switch.

Once you genuinely know what you are doing, speed stops adding information and starts adding damage. Now the work is delivery — how well, how consistently, how carefully, with what standard. That phase rewards meticulousness and punishes hurry, and everything good that will ever be said about the company gets decided there.

Go fast until you know what you are doing. Go slow once you do. Almost every company I have watched struggle had those exactly the wrong way round.

Slow discovery and fast delivery is the worst combination available. It is deliberating carefully in the dark, then shipping half-made things at scale once the lights come on. It is, unfortunately, what most well-resourced companies do — they can afford to be careful before they know anything and to be loud after they do.

Craft has its own clock

There is a third pace under both of these, and it is the one you cannot negotiate with.

Getting good at something takes the time it takes. Not calendar time — repetitions. You can compress the calendar by working harder, which I do. You cannot compress the repetitions. There is no arrangement under which a person becomes excellent at anything without doing it badly for a while first, and doing it badly is not a phase to be optimised away. It is the mechanism.

This is the part the modern conversation quietly refuses to accept, and it is why so much work now is technically competent and completely without judgment. The tools got faster. Judgment did not. Judgment is still built out of attempts, mistakes, and the slow accumulation of having seen how things turn out.

Which brings us to the machines

A new model lands every few weeks. The pace of the tooling is genuinely unlike anything before it, and it puts real pressure on the argument I am making.

My answer is a split, and it has served me well: learn fast, adopt slowly.

Learn fast — immediately, aggressively, the week it appears. Not knowing what these tools can do is now a form of professional negligence, and the gap between people who know and people who assume they know is widening every month.

Adopt slowly. Your systems are not a laboratory. The moment something becomes load-bearing — in how you deliver, how you make decisions, what a client depends on — it has to earn its place under the same standard as everything else you would put your name to. The speed of the tool is not the speed of your company. Confusing those is how organisations end up rebuilt every six months around whatever arrived most recently, which is not innovation. It is just churn with better press.

Trees whose shade we will never sit under

There is a saying I have loved for a long time: we are planting trees whose shade we will never sit under.

I believe it about this work in a way I did not expect to when I was younger. The point of building here is not the outcome I will personally get to enjoy. Somewhere ahead there is a young person who does not experience unemployment — not because of a policy, but because somebody built a company thirty years earlier that grew into the thing that employed them. They will never know the name. They should not have to.

That is what a cathedral pace actually means. Not slowness for its own sake, and not the romance of patience. It means being willing to work at a standard whose payoff arrives after you, which changes what you are willing to do this quarter in ways that a shorter horizon never can.

You cannot cheat the timeline. But you can be extremely fast about finding out what to build, ruthless about not committing before you know, and then patient — genuinely, structurally patient — about building it properly once you do.

Slow is smooth. Smooth is fast. It was never advice about pace.

It was advice about order.