Building a startup can be like building a bridge. Originally, when people needed a bridge, they’d take sturdy-looking materials and build one. Usually it would be either ridiculously overbuilt or it would fall down. Then the Romans worked out how to use segmented arches and they built hundreds of them, many still in use.
But by the 18th century, industrialization and spreading commerce created demand for new kinds of bridges in new places. With finance, ambition, and arrogance to spare, lots of newly designed bridges went up, leading to spectacular and gruesome collapses. In 1845 in England, for example, a whole village ran to a bridge over the Yarmouth River to watch a clown float by, riding a barrel pulled by geese. When they crowded at the edge of the bridge, the chains supporting the truss on the opposite side snapped, throwing them into the river and killing 79, mostly women and children.
It wasn’t until the mid 1800′s that engineers started to understand the principles of preloading and the metrics needed to anticipate stress on building materials. For the first time in 2000 years, you could design a new kind of bridge and know in advance that it would stay up. Not that we’re 100% certain a new bridge won’t still manage to collapse; there are still circumstances people don’t know how to account for, but our process sure seems more like engineering than art these days.
Technology startups are “new kinds of bridges.” Each one of them is different enough, needs to bridge a different enough gap, that the old templates aren’t useful. Serious practitioners no longer take the templates offered by business plans seriously, at least for truly early companies. In the start-up context, even the best thought-out business plan is like a work of science fiction, or a map of unexplored territory, neither believable as a promise nor useful as a tool.
Instead, the lean startup movement has begun to generate and test approaches to developing new companies that may actually work. They are turning the old methods on their head – instead of pretending they know the unknowable, lean startups make learning their central business. The exploratory techniques being tested and taught at Stanford, Georgia Tech, and now Harvard, and funded by angels (but not yet venture capitalists) may eventually evolve into a powerful, cohesive methodology for discovering customers, finding viable business models, and figuring out, in your unique case, what to do next. In other words, we may be beginning to understand the engineering principles of startups.
It’s an exciting time, filled with heroes and fools, powerful ideas and seductive dead ends. We are business partners in a seed fund, serial entrepreneurs, and one of us is a Distinguished Professor directing the new accelerator at Georgia Tech, Flashpoint. The purpose of this blog is to report, review, kibbitz, stick our necks out, and generally further this potentially transformative enterprise: startup engineering.
Merrick Furst & Matt Chanoff
What do you think?