In an earlier piece I mentioned a CEO at one of my earlier companies who once looked at a service running on my desktop and asked whether I could make it public, right then — as it was, that afternoon. I half-dismissed him at the time. I said it was a story for another time.
This is that story. Though it turns out to be less about him than about what it does to a person to spend years in a room with two people who refused to bend.
He could not read the code. He never pretended otherwise. What he could do — what he did, every single time something started to work — was figure out where to go and sell it.
The company had grown out of a mobile-phone business that went into decline almost as quickly as it had risen. The next thing he reached for failed, and not quietly. Then one category started to move, and while it was still moving he was already gone, opening the next one alone. One of those eventually became the thing the company was known for. Another never took at all. He was rarely in the present tense.
For a long time I read that as restlessness. Later I understood it was the same engine that runs me — the one that cannot sit still once a thing works, that is already asking what comes next before the current thing has cooled. He had simply pointed it at selling instead of building. The salesperson version of the same engine.
He spent on the technical side without flinching when it counted. We started on MySQL, outgrew it, moved to Oracle Enterprise Edition, and later put the whole thing on Exadata — none of which he understood, all of which he approved, because the engineers told him it was what the next stage needed. He gave vendors a hard time when they earned it. I remember one storage system sold to us on the promise that it scaled: you simply added another unit when you needed more room. When the day came to add one, the exact model had been discontinued, and the replacement would not sit alongside what we already had. We ended up rebuilding the whole configuration. He was not gentle about that, and he was right not to be.
What he never had was the inside of the machine. What he always had was the one question that mattered to him: what is the thing here that no one else has? The structure that reshaped itself instead of being rebuilt. The numbers that were live instead of a day stale. The parts that did not fall over under load. He could not have written a line of it. But he always knew which part was the sentence you could sell, and he kept that sentence in his head, ready.
None of it was something he was born with. I watched it get made. It came out of the venture that failed and the stretch where money was tight enough that the company nearly went under. He was not, when I first met him, a person you would have called a hard worker. The shortage changed him. By the time it was behind us he had taught himself to find the one true selling point in anything — and that skill was scar tissue, the same as anyone's.
The CFO held his line just as hard, in a quieter register. He carried risk that a job title does not capture — the kind you take on with your own name attached, when the company you believe in is closer to the edge than anyone outside the room can see. The specifics are his, not mine, and they are not going on this page. But I watched him keep the numbers that should have kept us all awake off our desks, and carry them himself — the company's survival was never someone else's problem to solve. It was his, in a way that cost him personally.
Neither of them started from a name, or from the shape of a thing. The work came first; the name, if it ever arrived, arrived after the work had earned one. Each held himself to producing a result at the execution level, with his own hands, and when something fell short, no one in that room reached for someone to blame. Each had his own way of operating. The details evolved over the years, but neither ever bent the core of it. Whether the way was right or wrong, in the end, mattered less to them than that they believed in it — and in themselves, doing it their own way.
There was a stretch, when the money was tight, when the three of us each did the most our own role allowed and nothing less. That is when the company turned.
I was the third one in that room. The CTO. And what being there did to me was not hand me a method. It gave me permission to have my own — or, closer to the truth, it made clear that I had no choice but to build one.
It is still most of what I run on. That one person can be a complete unit of responsibility, not a fraction of one. That the next thing should already be in mind before the current one is finished — I caught that from the CEO directly. That the result is mine to deliver, and when it breaks, mine to answer for, with no one else in the sentence — I caught that from both of them. None of it arrived as advice. It arrived as two people, in front of me, every day, refusing to do it any other way.
Whether I belonged in that room — whether the two of them would have called the three of us a team — I don't know. I used to want to know. I've stopped needing to. That part is two-sided, and I only get to speak for mine.
My side is simple. I was there. I learned this. And I have been grateful for it for a long time, in a way that doesn't depend on the answer to the other question.
Built with Claude (Opus).
Earlier in this series:
- The Accordion Pattern: Why I stopped writing one fat LLM prompt
- Nobody knows when a job will finish. I'd still like to report it accurately.
- What survives when you build alone for 24 years
- Dynamic isn't enough. Operations is the other half.
- Live report: at this speed, you don't theorize. You eliminate.
- The loop I didn't notice closing
- Abstractions are fine. Starting on them isn't.
- Twenty four years, ten DB migrations, zero downtime
- Write the code well once, the spec stops bothering you
- The 3-line discipline
- How I removed the middleman, one phone call at a time
- The graph nobody is watching
- I survived 24 years because I'm lazy
- Three failures I still think about
Top comments (0)