In 2001 I started writing software for entrepreneurs no one was serving. Twenty-five years later, the same product is still running — used by more than 250,000 entrepreneurs across 190 countries, by close to 20 million people in total. Most software companies that were started the year I started Business in a Box no longer exist. The ones that do, almost without exception, are the ones that figured out something durable about who they were serving and refused to drift.
What follows is the operating version of that story — the version that shapes the work we are doing now at Biztree Holdings. Not the polished one.
The problem we built into
The big productivity suites of the early 2000s assumed you had a back office. The web tools assumed you had a developer. There was almost nothing for the founder who needed a contract, an HR policy, a marketing plan, and a working business model — all today, none from a lawyer's hourly rate. That gap was enormous and the existing software market was patronizing it.
So we built into it. Business in a Box began as a library of documents and tools that gave entrepreneurs the operational scaffolding they could not afford to assemble themselves. It worked because the problem was real, the customers were underserved, and the product made them measurably more capable on day one.
What 250,000 customers actually taught me
- Customers tell you what is wrong, eventually. They rarely tell you on day one. They leave instead. The companies that learn to read silence as carefully as feedback are the ones that compound. Most of what we got right in the second decade came from listening to the patterns that had been there in year three.
- Distribution is a flywheel, not an event. We grew Business in a Box almost entirely through word of mouth and search. Paid acquisition was a multiplier, never a foundation. A product good enough to be referred grows without the ad-spend treadmill — and that compounds for decades, not quarters.
- Retention beats every other metric. I have seen a thousand pitches lead with growth, CAC, and ARR. I have seen very few founders make retention the first slide of the deck. The ones who do are usually the ones running real businesses.
- Discipline shows up in the P&L long before it shows up in the pitch. Companies that maintain real margins through the years build optionality. Companies that promise margins they don't yet have build narrative — and narrative does not survive a difficult quarter.
- The most loyal customers are the ones who feel ownership. Their loyalty is a function of being listened to, not of being marketed to. Every meaningful improvement to Business in a Box came from a customer who told us what they wished the product did and watched us actually do it.
- Time horizon is the strongest competitive advantage almost no one uses. A founder willing to hold for fifteen years operates against a market that mostly cannot see past three. That asymmetry is the entire game.
Why I did not sell
Most founders sell when the company reaches its first peak. I did not, for a reason that becomes obvious only with distance: the second twenty years of a good company are almost always more valuable than the first. The compounding takes time to show. By the time it is visible to the rest of the market, the founder has already been collecting it for a decade.
Holding Business in a Box through several technology cycles — the move to SaaS, the rise of mobile, the consumer-software professionalization wave, and now the AI rewrite — taught me something that is hard to learn any other way. Each cycle looked, from inside the cycle, like it required selling and starting over. None of them did. The product had to evolve. The thesis did not.
Why now is the largest opportunity I have seen
AI rewrites the unit economics of every category I have operated in. The work I have spent twenty-five years doing — building software that helps people run their businesses — is going to be redone, faster and more capably, across every adjacent space. That is not a threat to anyone who has been paying attention. It is the largest opportunity I have seen since I started.
The combination of AI and a quarter-century of operating reflexes lets a small group of operators build many companies simultaneously instead of one — and apply the same discipline to each that takes most founders a decade to develop. That is the work we are doing now at Biztree Holdings.
What this experience tells me about the next chapter
The companies that win the next twenty years will look like the ones that won the last twenty: durable customer love, real retention, a culture that ships, ownership that lasts. The shape of the work will change. The fundamentals will not.
What changes is the leverage. A holding company built around permanent capital, operator-led decisions, and AI-native architecture can compound across many companies the same way a single well-built product compounds across many customers. The discipline does not have to be reinvented at each new venture. The customers — the entrepreneur trying to start something, the team trying to do more with less, the person trying to live with more clarity — get more, faster.
I have spent twenty-five years doing this once. I plan to spend the next twenty-five doing it many times.