Business By Assembly
If a company needs more than two or three critical factors to succeed, the odds drop fast. Each variable you have to get right multiplies the difficulty. Three variables at 70% confidence each: 34% chance of success. Add a fourth: 24%. The math is unforgiving.
Most businesses have one or two genuinely novel elements. Everything else is commodity — solved problems with existing solutions. The novel pieces are what create the edge. The rest you can buy, license, or rent.
Identify which one or two variables actually move the needle. Assemble everything else. Not because it’s cheaper — often it costs more. Because your problem-solving capacity is finite, and spending it on commodity problems is the most expensive thing you can do.
The anti-pattern isn’t founders building everything themselves. It’s founders burning months solving problems someone else already solved — custom-building tools, reinventing operations, optimizing things that don’t matter — because solving commodity problems feels like progress.
They could pay more and get there faster. Instead they waste time saving dollars, not realizing the time costs more than whatever they saved.
The real cost isn’t money. It’s attention on the wrong problems.
When I started Ashore Hospitality, this is how we approached it. Short-term rental market is fairly efficient for properties in high-demand areas — that’s commodity. Booking platforms, pricing algorithms, automated guest communication — all existing tools. None of that was novel.
The novel insight was operational. Reviewing the properties’ management statements, I saw significant margin built into maintenance costs. We could improve profitability by cutting that margin and systematizing operations. Budget secured on day one to implement systems, focus on the operational edge, let everything else run on assembled components.
The edge wasn’t in the rental business. It was seeing that the cost structure was wrong and knowing we could fix it. One variable. Everything else was assembled.
Branson operates the same way. Launches, tests within two weeks, discards what fails, advances what works. Licenses, acquires, integrates wherever he sees advantage. The assembly model makes that speed possible — you can’t test a hypothesis in two weeks if you spent six months building infrastructure for commodity problems.
Assembly makes the bet cheap to make and fast to evaluate. If it fails, you haven’t sunk months into custom infrastructure. If it works, all your attention is already on the thing that matters.
The assembly isn’t the insight. The insight is knowing which piece to bet on.