People assumed two things. That something must have gone wrong. And that I would be going into consulting.

The second is easier, so I will start there. No. And the reason is the reason for all of it.

I had wanted to build something for years

Not recently. For years.

It sat in my head long enough that I stopped treating it as a plan and started treating it as something I thought about on planes.

I liked the work. I liked building teams that outran themselves. I liked walking into a problem nobody had named and naming it. I liked backing people early and watching them get further than they expected to — that one more than the rest.

And underneath all of it, the thing I could not put down. I wanted to build one of my own.

So the question was never whether. It was when.

The answer was not courage

I used to think I was waiting for nerve. I was not.

I was waiting until I had the time, the experience, the skills and the resilience to do it properly. And until I could afford to start it for the right reasons rather than the urgent ones.

Those are four different things and they arrive at different speeds. Skills came first. Experience took longer. Resilience took longest, and I only know I had it because of what the last six months have asked for.

You do not leave a good career because it stops being good. You leave when you are carrying enough to build something that holds.

Which means the twenty years were not what I escaped. They were what made it possible. Every operating model I inherited and had to fix. Every transformation I watched stall. Every restructure I sat inside. That is the raw material, and I did not walk away from it. I brought it.

The same law, somewhere else

They would buy the technology. They would hire the talent. Nothing would change. Then everyone would look for someone to blame.

It was almost never the technology. It was almost never the talent.

It was that the organisation was not built to carry what it had just bought.

Careers work the same way and almost nobody says so.

If you are stuck, you have probably decided it is because you are not good enough, not visible enough, or not lucky. It is usually none of those. It is that you are trying to carry something you have not yet built the structure for.

Or — and this is the more common one — you are already carrying plenty and have not noticed.

Effort is not the lever. Capacity is.

What I got known for

Over twenty years you acquire a reputation whether you manage it or not. Mine came down to a few lines, and I heard them often enough to know they were real.

If you need a difficult programme mobilised, give it to her.

If you want people who will be running things in ten years, put them in her team.

If you want radical candour, work with her.

And the fourth, which is the third said by someone who meant it differently.

If you do not want to hear the straight version, she is not for you.

That one cost me rooms. There were conversations I was not in, and I knew at the time which ones and why.

It was not always comfortable. It was not always good. But it was me, and I was never willing to trade it.

Which brings me back to the question

I had the equipment for a while before I moved.

That is the honest part. I had a list of reasons to wait and every one was reasonable. One more cycle. One more year. One more thing to see through. I could have defended any of them in a boardroom. I did, to myself, repeatedly.

Then my eight-year-old asked me why I had not done it yet.

I gave her the reasons. She listened to all of them and said: why not?

I have sat in rooms with boards. I have defended business cases I did not fully believe in, and won.

I did not have an answer for an eight-year-old.

She did not make me ready. I was already ready. What she took away was the need for permission — and there was nobody left to give it to me anyway.

Six months in

It is hard. Every day, and I am not going to pretend otherwise. Anyone who tells you the first year of building something is a relief has not built anything.

But the hard part is not the one I expected. It is not the money, or the hours, or the not knowing.

It is the unlearning.

Twenty years of muscle memory, and the strongest of it is this. I still catch myself waiting for a decision that is mine to make. Looking up for a sign-off from someone who is not there. The pause is only a second or two now. It used to be most of a day.

That reflex was an asset for twenty years. Checking, aligning, bringing people with you — that is most of what good operating leadership is, and I was good at it. It is also the most expensive habit to carry into something of your own, because there is nobody above you and the waiting is pure cost.

Relearning is the easy half. It is additive. It feels like progress and you can put it on a list.

Unlearning is subtractive. You put down the thing that got you here, and nothing replaces it for a while.

That is what this is called, and that is why.

And here is the part I did not plan for.

The thing that occasionally cost me rooms is now the thing people pay for. Boards do not need another adviser who will agree with them expensively. They need someone who will say the structural truth early enough to be useful.

I spent twenty years being told it was a strong flavour.

I have built a business where it is the service.

The advice I was given most often was to start with one thing and do it well. I did the opposite, and I would do it again.

Since I announced I was leaving, one question has come back more than any other. So what are you actually doing — consulting, or advisory? Neither, quite. Working out why took four months, and I was building the thing the whole time.

That last part is what I want to write about, because it is the part that gets left out. Every account of starting something has a chapter where the founder sits down and designs the business. I have read dozens of them. I have never met anyone who got that chapter.

You do not get a blank month. You get clients arriving before the process exists, a website going up while the positioning is still moving, and a decision about engineering that has to be made long before you know how much work there will be. The model, the operating system, the stack and the growth are all built at once, by the same person, in the same week.

That is not a failure of planning. It is the actual condition — and it is exactly the condition every growing business is in, which is why so many of them end up with seven good tools and no answers.

One decision did come before all of it. This was never going to be a consultancy only. Twenty years taught me to name a problem precisely, and naming it was never what I wanted to do. I wanted to be accountable for fixing it. Everything below is downstream of that.

Month one — what a blank page actually costs.

What is this actually going to be?

Three arms: Advisory, Products, Fluency. Four markets rather than one. What each arm sells and to whom. The pricing logic — not the prices, the logic. And the boundaries of the IP, drawn early, because they are almost impossible to draw later.

None of it was built. All of it was written down.

I want to be honest about how unglamorous that month was, because the temptation to skip it is enormous. There is nothing to show anyone. No product, no client, no revenue. Just a document that changes every second day.

But on paper, a business costs nothing to change. Everything after that has a price — in code, in commitments, in what you have already told people. Sense is the cheapest month you will ever have and the most consequential. It is also the one every founder I know wishes they had spent longer in.

Month two — advice, not feedback.

Does any of it hold outside my own head?

I took the model to a dozen people. Advisers, former colleagues, people who had sat in the seats I was describing. Not a survey — real conversations with people whose judgement I trust and who had no reason to be polite.

The three arms held. The four markets held.

The language did not. Three drafts of the positioning died in those rooms, and the version that survived came out of a conversation rather than a document.

Here is the thing I would tell any founder, and it took me most of that month to learn it.

Ask for advice, not feedback.

Feedback gets you encouragement. And encouragement kept coming — warmly, constantly, from people who genuinely wanted this to work. Almost none of it was information. It felt wonderful and it moved nothing.

Advice is different, because advice requires the other person to imagine themselves buying it. To put their own money or their own reputation next to the thing, and then tell you what stops them.

The question that finally worked was not what do you think of this. It was: what would stop you from buying this? The first time I asked it properly, the answer cost me a draft. That was the most useful conversation of the month.

Month three — the decision before the tool.

What has to exist for me to be accountable, not just right?

Month one’s conviction had a cost, and this was the month to pay it. Advice you cannot deliver is only a recommendation. So there would be engineers — not a vendor list, a team. One team, two countries, and no handoff between the people who design and the people who build.

Then the less visible work. Who holds judgement, brand and the client. Who holds architecture and build. How work actually moves — enquiry, conversation, proposal, delivery, invoice — written down before any software was involved.

I nearly bought software in month one. Buying things feels like progress, and it produces something you can look at on a Tuesday.

A tool cannot fix a decision nobody has made. Deciding who decides was the harder work, and it had to come first. It is also the single most common thing I watched large organisations get backwards for twenty years, usually at a cost of about two years each time.

Month four — going live is a stage.

What has to be running before anyone arrives?

The site. The rooms open for applications. The operating layer carrying real work rather than a spreadsheet and a good memory.

This was the first month where things built earlier started paying for themselves rather than costing time. That compounding is real, and it arrives later than you want it to.

Going live is a stage, not a deadline. It deserves its own design. Most people treat it as the moment the work stops. It is the moment a different kind of work starts.

The thing I could not buy.

Somewhere in month three I did what every growing business does. I went looking for tools.

I found them, and they were good. A proper CRM. Something for invoicing. Something for the project work. Something for content. Each one solved the problem it was built for, and solved it well.

Then I asked a question that crossed two of them.

How much of this month’s revenue came from conversations that started in the rooms? Which client costs more to deliver than it earns? Where is capacity actually going?

Every one of those needed three tabs, an export, and a spreadsheet I built by hand and never opened again.

The tools were answering me point to point. Nothing was joining them up to serve the business.

That is not a tooling problem. It is an operating system problem — and it is precisely the gap I spent twenty years watching large organisations paper over with headcount. They can afford to. A business of one cannot.

So the team and I built one. Not because we set out to make a product. Because I needed the thing to exist in order to run the business, and we happened to be the people who could build it.

It was never about the hours.

I could tell you how many admin hours came back. That is the number every product in this category leads with, and it is the least interesting thing about any of it.

What it gave me was awareness. One place that tells me what is happening, what is due, and what is coming next month. I open it in the morning and I know where the business is — without assembling it from four places and my own memory.

And awareness is what buys the decisions. Before, I made calls on instinct. Not because instinct was better, but because getting the information cost more than the decision was worth. Now the call gets made on something.

I have said for years that time is the only thing you cannot get more of. Running this on my own made that literal in a way it never was inside a large organisation. Every hour on admin was an hour losing money or losing time with my family. Usually both.

And the worst of it was never the hours. It was carrying a hundred small unfinished things around at nine o’clock at night. None of them important. All of them mine.

That is what went away. Not the tasks — the low-grade permanent worry that something has been dropped.

What did not work.

Three, honestly.

I mistook encouragement for validation. Everyone I spoke to in month two was supportive. It took several conversations to work out that supportive is not the same as would buy, and to start asking the question that actually hurt.

I over-configured before I had volume. Two modules set up for a business three times the size, then simplified back. Configure for the next six months, not the next three years. The version of you that needs the complexity will build it better than the version of you guessing at it now.

I wrote the process after the first client, not before. It should have been before. I now do that work in the first week of every engagement, because I know exactly what it costs not to.

The four questions.

If you take one thing from this, take these. One per month, in order. They are the whole method and they cost nothing.

1. What is this actually going to be?

2. Does any of it hold outside my own head?

3. What has to exist for me to be accountable, not just right?

4. What has to be running before anyone arrives?

If you cannot answer one and two, do not spend money on three.

That is the mistake I nearly made, and it is the one I see most often — in businesses of one and in businesses of ten thousand. The size changes. The mistake does not.