The piggyback problem – TechCrunch
I wanted to write about scooter startups this week, but, alas, I failed to care enough about them to muster any opinion at all. The problem is that they are pure piggyback startups, and pure piggyback startups are boring because they have no chance of being genuinely transformative.
Let me explain. Many, or even most, successful tech startups / movements succeed because they manage to piggyback on existing infrastructure. This is so painfully obvious it’s almost a truism, where the infrastructure is “the Internet” or “smartphones” — but there are other kinds, too. In its early days, Amazon was a pure piggyback startup, relying on UPS/FedEx/postal infrastructure. Similarly, the scooter startups are obviously reliant on existing city infrastructure.
Hollywood movies follow a three-act structure, and so do transformative tech startups and movements. Act I almost always consists of piggybacking on pre-existing infrastructure. In Act II, they build / evolve their own new, custom core infrastructure. And in Act III, their new platform begins to supplant and obsolete existing / establishment infrastructure.
Consider Amazon, who have evolved their own infrastructure in the form of gargantuan and increasingly automated fulfillment centers — Act II — and are now reportedly launching its own delivery service, while decimating shopping malls — Act III. (Though it’s true that the so-called “retail apocalypse” is more complicated than that. ) Consider Uber and Lyft, who are still in Act I, relying on externally driven vehicles, but fighting to transition into a self-driving Act II.
Amazon also combined Acts II and III with AWS, of course, since that was a once-in-a-generation case where there was no existing/establishment infrastructure. Similarly, Google has a long history of unleashing its internal Act II infrastructure to become Act III industry transformers, eg MapReduce, Kubernetes, and TensorFlow.
An even more unusual example is Bitcoin, which evolved its parallel infrastructure (miners and nodes) from scratch straight into Act III, an extraordinary instance of bootstrap levitation. This succeeded at first purely because it was so technically interesting and innovative, and subsequently because it was built from the ground up to incentivize infrastructure growth — to the Sorceror’s-Apprentice-esque point where it’s possible that as much as one in every thousand watts of electricity generated worldwide today, and counting, goes to securing the Bitcoin blockchain.
You usually want to piggyback before you evolve your own infrastructure, lest you become Webvan, although there are several spectacular exceptions. Elon Musk has spent his career trying to build third-act companies; PayPal topped out at Act II, so he went on to SpaceX (which started in Act II after Musk’s attempt to piggyback on Russian ICBMs didn’t work out, and is now clearly in Act III, beginning to supplant the existing launch-industrial complex) and Tesla (which similarly launched into Act II and, is extremely ambitiously, aiming for Act III vs. the multitrillion-dollar installed base of global oil infrastructure.)
But, like American lives, some startup have no second acts. This is what I call the “piggyback problem”; when there’s no apparent way to evolve your own infrastructure. To be clear, this is not necessarily a business or financial problem. AirBnB leaps to mind as an example of an extremely successful pure Act I startup; it’s made arguable attempts towards Act II infrastructure, but I don’t think the path there is particularly clear. I’m sure its founders and backers are weeping all the way to the bank.
However, this does make AirBnB a little … well … boring. And the same is true of scooter startups. They are all strictly Act I piggyback startups, and I can’t see how they might get to Act II in the viciously contested, heavily regulated environment of the modern city. (Lest anyone argue that they are infrastructure, this is only true in a trivial sense; the point is that they rely on external infrastructure.) Not that there’s anything wrong with being Act I. But Act III is where tomorrow is born.