Not Every Venture Is Trying to Win the Same Game
You don't have to have a company to start a company.
attributed variously in startup circles
I built an app with exactly one loyal user. Me.
Fitspan started as a longevity fitness tool I wanted for myself, vibe-coded over a few weekends after reading Peter Attia's Outlive and deciding I wanted something that actually tracked the metrics that mattered rather than the ones an app store algorithm thought I'd like. It worked. My best 2.4km time (as of Sept 2026) is now 10 minutes 9 seconds — significantly faster than when I started and comfortably meeting the fitness standard I had set for myself. I shared it with a few friends. Almost none of them stuck with it.
For a long time I would have called that a failure. I don't anymore, and figuring out why took looking honestly at four other ventures that didn't succeed in the same way either — because it turns out none of them were ever playing the same game.
Some ventures exist to solve your own problem
Fitspan was never built to be a company. It was built to be a tool, for an audience of one, that happened to also be shareable. The lesson here isn't about product-market fit — it's that the "market" was optional from the start. I got what I wanted out of it (a faster 2.4km, a clearer sense of my own longevity metrics) regardless of whether anyone else adopted it. Measuring it by user growth would be measuring the wrong thing entirely.
Some ventures exist because a capability is scarce
Postalcodes.sg was different — built explicitly for income, in the years before AI made its core skill common. It sold cleaned, structured address and geolocation datasets to small and medium enterprises that needed them for navigation and mailing. I built the entire pipeline — scraping, cleaning, structuring — in Excel macros and Python, standardized enough that a new dataset only took two to four hours of my own time every six months, on top of a fully automated ten-day scrape. It was, for a while, about as close to passive income as a data business gets.
It's tapering off now, and not because the market disappeared — because the skill I automated stopped being scarce. AI tools can now scrape and clean a comparable dataset without me at all, and the per-call API pricing model that used to work simply doesn't hold up anymore. I'm honestly not sure it's still a relevant business in its current form. But it taught me something worth more than the revenue did: a business built entirely on automating a tedious task has a shelf life exactly as long as that task stays tedious for everyone else.
Some ventures exist to prove you can just build the fix yourself
Long before any of my own companies, I watched a fellow military staff officer in 2009 build a simple SMS tool that notified senior leaders when a meeting's timing shifted — sparing them hours of waiting around as earlier agenda items ran long. It wasn't a business. It was a fix, built by someone who noticed a gap and had just enough technical skill to close it himself rather than waiting for someone else to.
That stuck with me long enough that I taught myself to code because of it. Years later, with tools like Power Apps and Power Automate lowering the bar further, I built a hospital discharge app, automated forms, and staff attendance logging for a previous employer — none of it spun out as a company, none of it needed to be. The value wasn't in owning equity. It was in being the kind of person who sees a gap and can just build the fix.
Some ventures exist to teach you what building a business actually requires
Hospitoolkit, Wikidoctors, and Privihealth all started the same way: I saw a real gap in healthcare delivery or operations, and I built something to address it. Each time, I underestimated how small a fraction of the work that actually was. Building the product turned out to be the easy part. Wikidoctors' hardest problem was never the platform — it was getting doctors to actually adopt it and contribute content, in a professional culture with little existing incentive to do either. The site has been on hold since 2025, not because the idea stopped being good, but because other priorities won out over the much harder, much slower work of building distribution from nothing.
That's the lesson these three ventures share, whatever ultimately happens to each of them: identifying the problem and building the product is often the smallest part of the journey. Getting people to change their behaviour around what you've built is where most of the work begins.
Some ventures exist to test whether you can trust the right people
Baliquarters is the newest kind of bet for me — a real estate development built with two friends I've known for more than thirty years, one an architect, the other a financier with real estate sales and short-term housing management experience. On paper, it looked like the obvious move: combine complementary skills, add my own expected tenure in Bali, and build something with long-term, largely passive value.
In practice, the first eighteen months were genuinely hard — real disagreements about where to spend limited capital, how far to push an ambitious design against what the target buyer actually wanted, how much risk any one of us was comfortable carrying. Construction is complete now; sales are still ongoing. The most valuable lessons so far have had little to do with real estate itself. It's a much better sense of how to actually source people you can trust with money and judgment, not just people you like — and an ongoing education in what makes a stay, or a sale, actually feel good to the person on the other end of it, not just profitable on paper.
Judging all five the same way was the actual mistake
Looking back, the mistake wasn't that some of these ventures grew slowly, stalled, or changed direction. The mistake was assuming they should all be judged by the same scorecard.
None delivered the same return. That doesn't mean they failed. It means they were investments in different things.
The older I get, the less convinced I am that every venture needs to become a large company. Sometimes the most valuable outcome is simply that it teaches you something you couldn't have learned any other way.
Looking back, I probably learned more from the ventures that struggled than from the ones that worked exactly as intended.