In 1872, something strange took place in a small town in Pennsylvania. A seemingly mad man of Scottish origin was constructing a mysterious plant.
It turned out that the man was named Andrew, and that he was planning to produce steel in huge quantities.
The strange part was not the steel part. Steel had been around for thousands of years. The strange part was his ambition to make so much of it. Because at the time, steel was mostly used by artisans for jewelry. And they did not need much.
But Andrew was convinced that steel had superior properties. He imagined that steel could support buildings tall enough to scrape the sky, and bridges long enough to cross mighty rivers. To most people, it sounded like science fiction.
However, Andrew did not just make steel. He perfected it. Andrew refined the processes of steelmaking and broke new grounds in quality and cost.
Still, he produced more steel than was needed, and things looked bleak. In response, Andrew bet his future on a single audacious project. One that would either prove he was right or utterly humiliate him.
In 1874 Andrew revealed the world’s longest bridge (of its kind). The first bridge to cross the massive Mississippi River. Built entirely with his steel.
The immediate reaction was disbelief. Everyone knew that nothing sizable could be made with steel. And certainly not a bridge. But Andrew fetched an elephant he had borrowed and crossed the bridge with the enormous animal. This inspired confidence and hordes of people followed while newspaper photographers secured the frontpage.
Following the opening of Eads bridge, steel became a critical enabler of the Industrialization. At the time of Andrew Carnegie’s death in 1919, American cities had been utterly transformed. Iconic skyscrapers towered over Chicago and New York. Science fiction, indeed.
The real value of steel
Telling the story of magnificent steel skyscrapers paints an illustrious picture. However, it does not do steel justice. In fact, skyscrapers were one of the lesser impacts of steel.
The real benefit of steel was for machinery. The properties of steel made it uniquely suited for tools and machines that entrepreneurs could use to produce a wide variety of new innovative products.
In the years following the adoption of steel, a famous cohort of entrepreneurs used steel machines to make: Proctor and Gamble soaps, Levi’s jeans, Ford cars, Edison light bulbs and Heinz ketchup. Just to mention a few.
In other words, steel-based machinery served as infrastructure to produce new products. And as we will see, the distinction between infrastructure and product is key when timing investments.
Infrastructure before products
Entrepreneurs have limited resources. For this reason, their ideas need a mature infrastructure. A crafty entrepreneur prior to 1872 might have envisioned a skyscraper. But before Carnegie steel, eighty story buildings were not possible.
More than a century later, Reed Hastings of Netflix also had to wait for broadband internet to mature before he could realize his vision of streaming (until then he had to make do with enveloped DVDs).
Unfortunately, it is hard to know when infrastructure is mature. Those who invested in electric cars in the 1980s, webshops in the 1990s, and mobile applications in the first half of the 2000s learned exactly how hard.
In these cases, it turned out the investors were too early. But, why exactly do investors lose money when being too early? Because the products never get good enough before the companies run out of money. And why don’t the products get good enough? Because the technologies powering the products are neither powerful nor cheap enough to serve as effective infrastructure.
Electric cars in the 1980s did not have the lithium-ion batteries and AI that power a Tesla. Webshops from the 1990s did not have the payment processing and high-resolution imaging that power Shopify. Mobile apps before 2007 did not have touch navigation and GPS that power Pokemon Go.
The entrepreneurs behind electric cars, webshops, and apps need the infrastructure to reach maturity. Or more precisely, they need the many individual pieces of the infrastructure to converge and reach maturity in unison. This is an important detail because the infrastructure of most innovative applications are a mix of many individual innovations. A fact so important, it warrants the naming of a law.
The law of compound innovation
The story of Carnegie left out an important detail.
As you have already surmised, steel was not the only innovation required for Henry Ford to produce a car. Nor was it true for soap, ketchup, and jeans. In fact, the infrastructures were a mix of different innovations that converged and matured in unison. Besides steel, simultaneous advances in electricity, gasoline, and rubber were essential for car producers.
When products are built on infrastructure consisting of multiple innovations, two things happen. First, the timing of the maturity of the infrastructure becomes harder to predict. Two, the products that can be built, become harder to imagine. And as complexity theory teaches us, this effect is exponential. One might call it: the law of compound innovation.
But why does the law affect investors? Because investors cannot be too late either. In fact, it is the very nature of venture capital to invest before everyone else sees the value. That was how Light Speed Venture Partners made 2345x return as the first investor in Snapchat.
Talking about Snapchat, let us apply the law of compound innovation. It is fair to say that no internet expert in the late 1990s had foreseen the disappearing picture sharing app. Why? Because Snapchat required more than the internet. In fact, it required several innovations to compound.
To create Snapchat the infrastructure had to become mature enough for a couple of youngsters with no budget to build the first version. It required 4G connectivity, high-resolution mobile cameras, and app store distribution.
And when did these underlying innovations converge and form the necessary infrastructure?
The app store came in 2008. High-resolution mobile cameras started appearing in 2010. 4G was rolled out globally during 2010. The result: Snapchat launched in 2011. And so did all its cousins: Line, Viber, and WeChat.
Projecting compound innovation. Timing the future.
Venture Capitalists bet on the future. And for investors, the future is synonymous with timing.
Consequently, investors must construct a thesis about the future. And not just about what will happen (we all know that). But when it will happen.
In order to construct a valid thesis on timing, one must first understand the innovations that are forming new infrastructure. And the law of compound innovation hints that this becomes exponentially harder the more complex the infrastructure is.
Consider an emerging infrastructure like VR. In 2012, Oculus revived the forgotten dream of virtual reality. Almost eight years have passed, and very few people use VR. To understand why one must first understand the infrastructure for virtual 3D immersion.
In order to deliver a quality experience, one could theorize that the following infrastructure is needed: Wireless lightweight headset with long battery life and a screen resolution of 8K with 180 degrees field of view. 5G to stream the content. And controllers with individual finger and joint sensors. All within a price point of a mid-range smartphone.
In this light, it has clearly been too early for investors (and entrepreneurs) to bet on products like VR games, software, and films. Instead, investors should have been focusing on pieces of the infrastructure. Like controller and screen technology. (A topic for a later post).
But the thesis also hints that successful VR products could be close. The Oculus Quest headset from 2019 is not too far from the headset described. And 5G is being rolled out this year.
Consider a more complex infrastructure. The convergence of Blockchain and VR. An infrastructure that could be called “Virtual Society”. Blockchain is an infrastructure that allows to track and manage ownership. For Blockchain infrastructure to reach maturity, one could theorize that it needs: the ability to handle millions of transactions per minute, wallets to be pre-installed in browsers, and non-technical vocabulary for normal people understand it.
When VR and Blockchain mature and converge, the combined infrastructure will lay outside our experience. Much like the convergence of 4G, high-resolution cameras, and app store distribution was an unprecedented infrastructure that gave birth to the equally alien apps.
Perhaps we are already witnessing the first of such products. In February 2020, Decentraland launched. It is an immersive social network with blockchain-based ownership. An early example of products based on the convergence of VR and Blockchain infrastructure.
It is fair to say that Decentraland would be very hard to imagine a decade ago. But looking at the infrastructure powering it, you might be wondering if it is too early. And if you are, you are asking exactly the right question.
Perhaps the law of compound innovation can help you. See my first crude attempt at depicting it below: