When to build 0 to 1

Buy, build, or build software?

Real world businesses, long tail industries, and niche verticals are all the rage these days. So when should you build vs buy vs just build software?

We are very keen on funding software companies to buy and transform operating companies as opposed to selling software to them. Metropolis and Teamshares are our highest profile case studies. This approach works because selling vertical software is often too hard for too little payoff (low value capture in small markets) and building services companies from scratch usually sucks for venture (too many things have to go right).

Most of the ideas we see for GBOs (Growth Buyouts - buy and transform) or roll up platforms (be the best buyer) revolve around back office, CX, marketing, etc automation. They rarely touch “production”.

If you’re innovating on actual production (the work) vs tech that changes business operations (the work around the work), it might make sense to go full stack organically by building a service 0 to 1. It’s too hard to buy and refactor the core of the business around your new technology/innovation, especially when it relies on hard assets like machines, robots, architecture, etc.

A tech enabled buyout platform that fully refactors the core production will always be overpaying for capacity you not only don’t value but see as a NPV negative; in your hands the existing service is worse than worthless.

The obvious caveat is that if you’re able to buy at low multiple bc it’s “just a book of business” you can still justify buying and transforming the fundamental unit of work/service. You’re buying customers, not processes/talent/operations/assets and you can just throw all those away.

Obviously the underwriting criteria (for our money and your time) is more complicated and detail but this framework should at least help you select which approach to evaluate.