Reasoning from first principles is a sometimes brutal but absolutely essential tool that should be applied to the foundational idea of every technical startup. Because there are some things in life that aren’t merely difficult. Some things are impossible.
What makes a startup business succeed? That is the million-dollar question. However, “what makes a technology startup fail?” is perhaps an easier question to answer. This question, for the most part, would be answerable in the same way for any new business — technology startup or otherwise. There are hundreds of possible answers, so many you could write a book about it — and hundreds have.
I want to have a look at just one seemingly obvious answer to that question and ask “why is so often overlooked?” The surprising reality is that technology ideas (and technology startups) sometimes fail because the core idea just fundamentally won’t work — not because idea is badly executed, and not because there are confounding reasons or market circumstances. Rather, the basic technological idea just cannot work.
Is it possible for millions (or billions) to be invested in technology ideas that could and should’ve been flagged as a turkey from day one? You bet! Technology startup history is littered with examples.
A quick search today through any crowd-funding website will unearth dozens of over-subscribed and seemingly exciting projects that appear to solve real world problems with brilliant technology innovations. Roadways with built in solar panels! Crystal clear water from dry desert air! Chemical analysis of foods with your smart-phone that can detect pesticides and tell you when an avocado is ripe! Devices that plug into your wall and dramatically reduce power bills. Many of these crowd-funded startups are backed or endorsed by, various government agencies, universities and PhD’s. All with “working” prototypes and all just needing a bit more funding to get them across the commercial line. And yet, all doomed (inevitably) to fail.
Which idea would you back, given the chance? Any of the above, or some guy you’ve never heard of that tells you he can build a space rocket that will eventually bring the European and Russian space programs to their commercial knees? (Tip — go with the rocket guy).
Hindsight is great, but the real commercial trick is predicting the future — and for that you need to steal a secret from the rocket man himself, Mr Elon Musk. It’s a method of reasoning called “first principles analysis”. The principle is illustrated beautifully from a foundational pillar of his now hugely successful new-space startup, SpaceX.
When Musk returned empty handed from his trip to Russia trying to buy a slightly used ICBM for his Martian space shot project, it wasn’t his economics degree that gave him the path forward, or his software engineering credentials. Rather, it was his training in and absolute conviction in applied physics and first principles reasoning.
Musk asked himself and reasoned and researched a fundamentally simple question about rockets. What makes a rocket? Suppose you take a fully working (and fuelled) rocket and broke it down into it’s constituent parts — all of the aluminium, titanium, kerosene fuel, circuit boards etc and spread those parts out as raw materials and then added up the cost of those materials. What would that cost be? The perhaps astonishing answer is “not very much”, relatively speaking. Something like 1% of the then going rate for the fully working assembled and fuelled rocket.
It literally does not take a rocket scientist to figure out that if you can build other similarly complex technological systems at much less than a one to ten times premium over the cost of the raw materials (think of a car or computer or TV), it be surely should be possible for a rocket. And it turns out, it is.
Musk reasoned from first principles that there was no fundamental reason why he could not, with his modest fortune at the time and with the right expertise and approach, convert a few million dollars worth of readily available raw materials into a fully working rocket and thus dramatically reduce the cost of launching mass into space.
Building a commercially successful rocket is undoubtedly hard. Ridiculously so. Landing and reusing it every other week these days — triply so. But Musk’s reasoning established from a first principles basis that his rocket idea didn’t violate any basic laws of physics. It was, he figured, somewhat improbable that he could succeed at the execution, but it wasn’t actually impossible to do. It’s merely a very challenging technical problem that can be solved to reap great rewards.
So what about the water from air device? Or the chemical food analyser? Or the solar roadway? All ideas with obvious and huge commercial potential if they work. Idea plus raw material plus clever engineering and commercial savvy surely leads to commercial success? No.
A trained scientist or engineer can (or should) be able to look at any of the other above ideas and work out that, for example, a road is possibly the worst possible place to put a solar panel. A simple first principles analysis will quickly establish that cost ratio of solar cells to supporting (encapsulation) components to build a solar road overwhelmingly favours just placing the solar cell in a much simpler and cheaper enclosure on the abundantly available roof spaces. There are plenty of other issues too, but on the basis of that simple analysis, the idea is always going to be dead in the water, no matter how convincing the other arguments in favour might seem to be.
Similarly, we can calculate that the thermodynamics of a water-from-air system violates known physics when attempting to deliver anything close to the claimed benefits. Or that infrared lasers (used to perform chemical analysis of supermarket avocados) don’t penetrate through the skin of avocados (or anything much) and besides, are only useful to analyse very simple mixtures of one or two chemical molecules before the signal mixing becomes an exponentially difficult problem.
The idea might sound good. There might be some scientific principle that we’ve all heard of in secondary school (or university for that matter) that seems plausible. But for each of those failed ideas, there is a fundamental scientific gap that separates the idea from potential realisation. The problem isn’t merely difficult. It’s impossible.
Yet all too often the brilliant minds at work on the idea are so enamoured by promise of the end goal, or so focused on cleverly solving the technical challenges to get there that they, and their investors, fail to see the fundamental flaw in the underlying physics.
Taking a step back and reasoning out the feasibility of a technical idea is surprisingly hard to do and surprisingly often overlooked. Millions of dollars have been thrown eagerly at ideas that violate basic physical laws. New technology is exciting, especially if the idea has a compelling commercial case.
Weeding out the impossible from the possible is an essential first step. Of course, establishing that something could be achieved is also no guarantee of success. The execution has to be right as well. What could possibly go wrong with that? …
This post is the first in a series from EdgeGroup Incubator Founder Paul Schlusser.