Two Continents, and What Each Business Taught Me

I have founded and run businesses on both sides of the Atlantic — in Europe and in the United States — before and alongside the company I lead today. People usually ask which was the most successful. The more useful question is what each one cost me to learn, because that is the part that compounds.

What follows is not a case study. It is what I now know that I could not have been told.

Early on: I confused activity with progress

In the beginning almost everything feels like work, and almost none of it is. Registering, naming, designing, choosing tools, building the thing that will one day be shown to somebody. Every one of those tasks produces the sensation of momentum, and every one of them can be completed without a single person ever agreeing to pay you.

The uncomfortable discovery is that the sensation is indistinguishable from real progress while you are inside it. It only separates later, when you look back and count how many weeks passed before anything was tested against a stranger.

I no longer trust the feeling of being busy as evidence of anything. The only question that separates the two is whether the week produced information you did not have on Monday. Some of the most productive weeks I have had were quiet ones with a single hard conversation in them.

Then: I learned what a customer actually is

A customer is not somebody who likes what you are doing. Encouragement is free, and people are generous with it, particularly to someone brave enough to start.

A customer is somebody who changes their behaviour. They pay, they turn up, they rearrange something in their week to accommodate you. Until behaviour changes, all you have is goodwill, and goodwill is a very pleasant thing to mistake for a market.

This is the single most expensive lesson available to a founder, and almost nobody escapes paying for it at least once. I did not.

Now: I do not start what I cannot staff

It is different now, not because I became cleverer but because I finally understood what I was actually committing to.

Starting something is a decision that lasts one afternoon. Running it is a decision that recurs every week for years — the filings, the customers, the standards, the person who has to be answerable for it when I am somewhere else. Almost every venture I have seen fail was capable of being started and never capable of being run.

So the test I apply now is not "is this a good idea". Good ideas are abundant and cheap. The test is whether it can be operated with what already exists: the same standards, the same entity, the same way of working. If it needs a whole new apparatus around it, it is not an opportunity — it is a second job wearing an opportunity's clothes. I have written separately about why I refuse more than I build.

What the money teaches that a salary cannot

I spent years in advisory work, and I was well trained. But there is a specific kind of attention that only appears when the downside is yours, and no amount of professional diligence reproduces it.

It is not fear, and it does not make you timid. It makes you concrete. You stop asking whether a plan is sound and start asking who does it on a Tuesday, what breaks first, and what it costs to undo. Those questions are unglamorous and they are the whole job.

It also makes you far more forgiving of other people's operational decisions. Once you have chosen the imperfect supplier because the perfect one could not start until March, you stop assuming that odd-looking choices are stupid. Usually they are a constraint you cannot see.

What I would say before the first one

• Decide what would make you stop. Write it down while you are still calm. Nobody can define failure honestly in the middle of it.
• Find the first person who changes their behaviour. Before the brand, before the site, before the tools. Everything else is rearranging.
• Separate reversible from irreversible. Take a week over the second kind and an afternoon over the first. Treating them alike is the most common way founders exhaust themselves.
• Assume you will be the one running it in three years. If that thought is unbearable, do not start — no amount of early enthusiasm survives that arithmetic.

I am wary of anyone who claims a formula from their own handful of ventures, and I am not offering one. But it was enough to teach me that the hard part was never the idea, and never the starting. It was everything that happens on the ordinary mornings afterwards.

What none of them taught me until I looked back

Some lessons are not available in the moment. They only appear when you can look back across all of them at once, and they are the ones I would most want to pass on.

The first is that I was consistently wrong about which problem would end things. Every time, I worried about the visible risk — competition, demand, whether people would understand it — and every time the actual difficulty came from somewhere ordinary: capacity, attention, a dependency I had not registered as a dependency. The dramatic risks announce themselves and are therefore planned for. The boring ones are what get you.

The second is that I never once regretted a decision to be more careful about an irreversible commitment. Not once, in any of them. I have frequently regretted moving too slowly on something reversible, and I have never regretted taking an extra week over a signature. That asymmetry is so consistent that I now treat it as a rule rather than a preference.

The third is that the emotional cost of ending something well is much lower than the emotional cost of letting it drift. A clean stop is a bad week. A slow decline is two years of low-grade guilt while the thing consumes attention and produces nothing. I did not believe that until I had experienced both.

The question I now ask before anything gets started

There is one question that would have saved me the most, and it is not about the market or the model. It is: what has to be true for this to still be here in three years?

Not what has to be true for it to launch. Launching is comparatively easy and almost everything launches. What has to be true for it to persist — the demand recurring rather than being a moment, the supply not depending on one person's goodwill, the economics working at a boring volume rather than an optimistic one, and somebody being willing to run it who is not me.

Writing that list takes an afternoon. Every item on it is checkable, and most ideas fail two or three of them immediately. The reason founders skip the exercise is not laziness. It is that the answers arrive while the enthusiasm is still high, and nobody wants to look at them then.

On partners, briefly and carefully

I will not discuss any specific arrangement, because that would be neither fair nor useful. But the general lesson is worth stating, because it is the one people most often learn expensively.

Partnerships are almost never damaged by the thing that was written down. They are damaged by what was assumed — how much time each person would actually put in, what happens if one of you loses interest, who decides when you disagree, and what the arrangement looks like if the business succeeds far more, or far less, than either of you expected.

All of those are comfortable to discuss at the beginning, when everyone is optimistic and generous, and almost impossible to discuss afterwards, when the answer has become expensive. The discomfort of that conversation at the start is the cheapest insurance in business.

The version I now use is simple: before anything is agreed, both parties write down separately what they expect to contribute, what they expect to receive, and what would make them want out. Then compare. If the two documents match, you have a partner. If they do not, you have found out for the price of an afternoon.