The returns to innovation are wildly skewed. In practice, this means that the value of the vast majority of new ideas is small, even if they are executed correctly. But some will hit massively big, with huge payoffs. The outliers matter – a lot.
You can make a living with both kinds of innovation. For many years, Toyota executed thousands of small improvements to their manufacturing processes. Cumulatively, all of these changes added up to an entirely new way to build cars – even though there was not one single huge breakthrough included.
Home runs can pay off too, and when someone does hit one, we hear all about it. But usually, we only hear about it after the fact – once the innovation is huge. While it’s happening, the home run can look a whole lot like a strikeout. Just think about all of the grief that Amazon was given throughout the late 90s and early 00s.
The key issue here is this: if we are doing something genuinely innovative, we don’t know in advance if it will work or not.
One of the things that has bothered me in the recent discussion of “disruption theory” is the idea that it is a predictive theory. It’s not, and it can’t be.
If we are predicting innovations, we might be able to tell which industries could be disrupted. Chunka Mui and Paul Carroll take a good shot at this in The New Killer Apps. Deloitte Australia takes a crack at it as well in a report called Short Fuse, Big Bang. The folks at Pollenizer have been analyzing each of the industries featured in the Deloitte report – the most recent looks at finance. Here is the diagram that Deloitte has made:
So let’s say that we can agree on the industries that do face disruption – that these can be predicted with some degree of confidence.
We still face two problems.
The first is timing. People started talking about how the internet would disrupt journalism in the early 1990s. The full impact of this didn’t become fully apparent until about 2008. There were a lot of failed experiments in that time, but then we started to see some home runs, like politico.com and Huffington Post.
The second problem is figuring out which specific innovations will be the breakthrough ones. This is where things get especially hard. This problem has two parts to it as well. First, the past gives us poor guidance. Second, the only new ideas that will work are non-obvious when they start.
Nassim Taleb addresses the first issue in a must-read post on the limit of statistics. He outlines the types of payoffs that we can expect when we bet on the future:
Innovation (especially breakthrough innovation) lives in Taleb’s Fourth Quadrant. The payoffs are skewed (most returns are small, and the big wins are MUCH bigger than we would expect), and the payoff are complex.
Tren Griffin draws out what this means in a nice series of posts. He is mostly concerned with the Venture Capital angle on innovation, but the main points apply to anyone aiming for a disruptive innovation.
First, he quotes Mark Andreessen (Andreesseen is in bold, Griffin in normal font):
“The entire art of venture capital in our view is the big breakthrough for ideas. The nature of the big idea is that they are not that predictable.”
“Most of the big breakthrough technologies/companies seem crazy at first: PCs, the internet, Bitcoin, Airbnb, Uber, 140 characters.. It has to be a radical product. It has to be something where, when people look at it, at first they say, ‘I don’t get it, I don’t understand it. I think it’s too weird, I think it’s too unusual.’” This set of quotations reminds me of Howard Marks, who has said: “To achieve superior investment results, your insight into value has to be superior. Thus you must learn things others don’t, see things differently or do a better job of analyzing them – ideally all three.” The power laws in venture capital virtually guarantee that the poseur venture capitalist who follows the crowd can never make up for all their losers, since they will not get the one or two tape measure home runs required to generate returns that limited partners demand. What is perfectly advisable for the ordinary investor (“be the market”) spells doom for the venture capitalist because venture capital returns reflect the power law noted above. It is only in the non-consensus quadrants that optionality will be mis-priced and bargains found. Buying optionality is not enough to achieve success as a venture capitalist; it must be mis-priced. Paying too much of any asset including optionality is not a solvable problem. The matrix Marc Andreessen describes above, with an example in each quadrant, looks like this:
That diagram brings up the final issue here – that new breakthrough ideas are non-obvious. Paul Graham puts it like this:
Whereas if you want to start a startup, you’re probably going to have to think of something fairly novel. A startup has to make something it can deliver to a large market, and ideas of that type are so valuable that all the obvious ones are already taken.
Let’s sum up. Breakthroughs are where the big payoffs to innovation lie. However, while we might suspect which industries will see breakthrough innovations occur, it is impossible to predict in advance which ones will be winners. This is due to two reasons. First, statistics based on the past can’t guide us. Second, the ideas that win always seem a bit nuts at first.
Given all of this, what should we do?
First off, build learning loops into all of your operations. Here is Griffin again, this time discussing Vinod Khosla:
“Things go wrong. There is lots of uncertainty, and there are times when you’re unsure of yourself. I’ve found that the less people know, the more sure they are.” This is something that has always troubled me. A great story half told is somehow more convincing usually because muppets suspend disbelief since they want so very much to get rich quick. The best venture capitalist and founders are learning machines, because they realize that there is no end to what you can learn.
Second, take advantage of what Taleb calls optionality – Optionality is the property of asymmetric upside (preferably unlimited) with correspondingly limited downside (preferably tiny). Here is Griffin’s take on optionality:
“If you ‘have optionality,’ you don’t have much need for what is commonly called intelligence, knowledge, insight, skills, and these complicated things that take place in our brain cells. For you don’t have to be right that often. All you need is the wisdom to not do unintelligent things to hurt yourself (some acts of omission) and recognize favorable outcomes when they occur. (The key is that your assessment doesn’t need to be made beforehand, only after the outcome.)” Being able to make decisions which do not require correctly forecasting the future is a wonderful thing. Not one of the great value investors identified in the series of posts in this blog relies on macro forecasts of the future. Instead, value investors use the optionality of cash to buy after the outcome exists (i.e., a significant drop in intrinsic value). Regarding venture capital, Warren Buffett believes: “If significant risk exists in a single transaction, overall risk should be reduced by making that purchase one of many mutually- independent commitments. Thus, you may consciously purchase a risky investment – one that indeed has a significant possibility of causing loss or injury – if you believe that your gain, weighted for probabilities, considerably exceeds your loss, comparably weighted, and if you can commit to a number of similar, but unrelated opportunities. Most venture capitalists employ this strategy.”
This means that if you’re a firm, you need to be using the venture capitalist’s portfolio approach to innovation. Lots of little bets with big upside will beat trying to predict the future.
If you’re a startup, get into one of the industries that provides opportunity, find a non-obvious idea, and go after it as hard as you can. There’s no portfolio approach here, but your upside is enormous.
In both cases, we can’t predict success, we’re just making bets. We can improve the odds, but they’re still bets.