In our Growth Innovators Series, we’ve frequently mentioned to “The Law of Computability”. And we want to take some time to explore our definition and its implications for competition and why you should care about it now!
What is Computability?
A task is computable if it has two characteristics:
- It is highly digitized.
- There is a high level of knowledge about it.
The Law of Computability organizes this in a function: Computability is the product of the Degree of Digitization for a given phenomenon, and the Level of Knowledge about that phenomenon.
Generally speaking, the more computable a process is, the more efficiently and effectively we can execute it. For an organization, this leads to safer workplaces, more consistent results, higher margins, fewer employees per dollar of revenue, and improved earnings for investors. In many cases, it leads to winner take all markets obliterating the fragmented old-style competition. Google tore out the economic heart of the classified ad industry and created an almost monopolistic platform because it can compute desire.
In an increasing number of cases, having high Computability enables things that previously were impossible. Self-driving cars are only available because we have very high knowledge of the physics of driving, and now have cameras, LiDaR and other sensors that can sufficiently digitize the experience.
If we think about the industrial revolution as a redesign of life through new access to energy and the application of scientific management to drive massive automation and productivity, computability brings with it an even more comprehensive automation of the world. Its task is birthing a fully formed symbolic description of the target reality that springs from the computation process with the drama and possibility space that occurred when Athena sprung fully formed from the head of Zeus! And, computability of reality is happening faster, all the time... as Ray Kurzweil pointed out in his Law of Accelerating Returns, technology amplifies knowledge and knowledge amplifies technology in a positive competitive loop that drives capability and competition at higher and higher speed. It is the self-amplification process that makes things improve at exponential rates — and we are using "exponential" formally — not just to mean, really fast.
Is Everything Computable?
More and more domains fall before our ability to compute them the way that Medieval castles fell before the high compression cannon. Given our definition, it’s relatively easy to see that some industries are more “computable” than others.
Manufacturing, for example is highly computable. The entire business model of manufacturing implies very well-defined, highly repeatable tasks. So our “knowledge” of said tasks is incredibly high. And many of those tasks can be partially or sometimes fully digitized. We can track how it’s performing, all the variables at work in completing a given step, facilitating much more detailed reporting, predictive maintenance, and in some cases full automation.
Much of healthcare, by contrast, continues to be low on Computability. While we have a variety of medical technologies that allows us to digitize and measure various vitals, the human body is an incredibly complex system, and interventions have wildly variable results from person to person. Our knowledge is still frustratingly low. And while the bleeding edge of truly personalized medicine is coming (powered by sensors allowing us to digitize our blood glucose, our resting heart rate, our oxygen efficiency, our sleep quality and more), we’re not there yet.
This gets to a common misunderstanding between computability and automation. The major distinction between is the degree to which you can create a symbolic description of the task.
Automated tasks require knowledge, but not necessarily digitization. Automation has been with us for centuries, even though those solutions were entirely analog. They had no symbolic description. In many cases that was enough. But some innovations are simply non-starters without the ability to make that translation into symbolic descriptions, a prerequisite for digitization.
How to Increase your Computability
An organization looking to increase its level of Computability has three primary tasks:
- They need to understand what they know how to do, what processes they engage in on a consistent basis, and have those processes sufficiently documented.
- They need to take that knowledge and look for as many ways to digitize it as possible.
- Leaders must be willing to capture, buy, clean and invent new data to allow for faster digitization of the value chain.
Most organizations thing they’ve got the first step covered, but in practice many organizations are shockingly ignorant on how they actually do what they do. This is particularly true in professional service businesses, or businesses requiring a high level of person-to-person interaction and empathy.
In terms of the second step, this requires an organization to understand how investments in digitization pay off. The up front investment is certainly time, people, and capital intensive. You’re working to digitize a process, train the organization on how to use that newly digitized process, and engage in iteration to make sure it works as well as possible.
The third step often takes practical imagination. United Rentals placed sensors and communications devices on almost all of its assets, so it can know not only where any rented piece of equipment is, but is it running, does it need maintenance, etc. Any asset which is not tagged is a dead asset in a live world.
Overall these three steps can take budget: both capital investment and incremental labor. However, when it works, almost always leads to improved marginal and total economics. In the best case, it leads to brand-new assets in the form of proprietary data sets, allowing you to serve customers better, create self-perpetuating flywheels for growth, and developing processes that are difficult or impossible for competitors to copy.
The benefit of being an early adopter
Since 50% of the Computability function is one's ability to digitize a process, one can more easily see the upside of becoming an organization that performs strategic experimentation. New technology is continually coming out that unlocks new possibilities previously considered impossible. And yet most organizations look at emerging technology with skepticism, waiting for someone else to figure out how to apply it to their specific use cases.
Since Computability creates better products, better services, sustainable differentiation and more defensibility, it seems foolish to sit on the sidelines. And yet that’s exactly what most organizations are doing.
What’s currently unfolding in the realm of decentralized finance is a salient example. The existing financial world is full of intermediaries, each taking a cut of a transaction and in many cases creating a worse experience for customers. DeFi eliminates these middlemen, allowing new protocols and services to pass those benefits on to customers and to provide clearing services instantaneously.
Most importantly, by allowing each facet of a transaction to be computed, it enables vastly lower transaction costs — which can open up new types of contracts, payment mechanisms, insurance and credit. In many businesses (think about everything from layaway plans to auto purchases) the payment terms can differentiate your product or service — and blockchain could enable new value for customers.
How to get started
One of the tools we use at Manifold is the Computability Map. We work with an organization to define the scope of their knowledge - the most important processes that go on within an operating unit, division or organization - and help them assess how digitizable they are, factoring in cost, time and risk.
The result is a roadmap for how the organization or unit can increase their Computability, in what order. It’s an incredibly practical and helpful tool for organizations to get the lay of the land and have confidence in knowing where to start.
We often assemble an expert team to provide what we call an Expert Computability Review. You describe the nature of the challenge and/or task and a panel of our computability experts provide an outside in view on where your risks and opportunities might lie — which can be done in a week's time.
There are two reasons to start now.
One is a threat. If you have a significant digital competitor in your space, you must master computability to compete. (Ask Macy's how it is to fight with Amazon.)
Two is the desire to dominate. If you are the leader at computability in your space, it is probably true that your industry will consolidate and you can be the leader of that new, more concentrated market. There's a startup called ICwhatUC, in the HVAC repair space. They are starting as a service to help the homeowner self-repair or call a professional, but on the back of that business they are building a massive database on what the heating and cooling systems are in American homes and what state they are in. At scale, they will be more powerful than any other retailer/servicer in that space because they will be able to compute their customers needs.
As the old saying goes: the best time to plant a tree is 20 years ago, the second best time is today!