Abstraction vs. Disruption

By now we are all tired of hearing about ‘disruption’. The word has become over-used in many ways, and I am as guilty of its usage as anyone. There is something in writing about technology that calls for use of the word. In reality, not every new company is disrupting its industry. Instead, I think we are searching for a word to describe a specific phenomenon or fact pattern that we regularly encounter. Some companies are truly disrupting their industries, and a far smaller group may actually disrupt something bigger in the world. But for most companies there is another force at play, and we lack the right way to describe it. Since disruption sounds exciting we revert to that when in reality we are looking for a different term.

The root of the problem is that tech companies face a continual threat of obsolescence. This force has been at play for a long time going back to the first chip companies and later computer makers like Osborne which obsoleted itself to death. More recently there was MySpace and Friendster. In fact, I have been using the word “Tech” to describe what is really several industries – Semiconductors, Hardware, Software, and Internet. These are all very different but I think that in this case they share this common threat – if they miss the next product cycle they risk being permanently sidelined. Often, when we talk about disruption we are referring to this risk. My sense is that companies that make steel or cereal are far less exposed to this risk.

Understanding the risk of obsolescence is one of the keys to understanding technology. In business and management schools they teach the concept of Porter’s Five Forces. This names the pressures on a company as competitors, customers, complements, substitutes and suppliers. I believe that in technology there is one more force at play – abstraction. This is the idea that one task (or competitive strength) can be replaced with two tasks that are separate or abstract from each other. The most common way to think of this is software versus hardware. PC hardware makers do not have to build word processing software, instead they have the abstraction of an operating system (OS). The hardware maker builds hardware that supports the OS, and software makers build software that runs on the OS. And the two do not have to care about what the other is doing.

The notion of abstraction is a key element of computer science, one that is well understood by the engineers building products, but it is less appreciated by other groups making decisions around technology. In particular, investors have a hard time fully grasping this.

I see this play out frequently. A company that is currently the leader in one field wakes up one day to find that it customers have devalued half of its product by abstracting something away. For example, Cisco is the global leader in networking equipment. Recently, some of its largest telco customers have stopped buying Cisco hardware and instead insist that Cisco port its switching software to commodity hardware. Over the past 20 years Cisco has nailed down its position in the network, and monetizes this through the sale of high-margin hardware, where much of the competitive advantage rests. Suddenly they are being asked to break apart their bundle. Cisco can probably find a way to survive (and thrive) with this, but for many others this process can wreck a once healthy business.

This is separate from being replaced by a substitute. Abstraction changes the underlying way a a company conducts its business. And this is where you can see why it is tempting to describe the process as disruption.If you sit down and think about it in these terms you can probably come up with a half dozen examples easily. Google abstracted search away from Internet indexing, and it turned out that consumers wanted search far more than the indexing that Yahoo had previously provided.

Set against this force is the opposite one of bundling or integration. Apple being the obvious example of a company that integrates many previously discrete functions. In reality, abstraction and integration often go hand in hand, with one company or segment benefitting because of abstraction that takes place at a layer above or below them.

In semis and hardware, we often have to contend with the “Good Enough” threshold. Here, companies that sell discrete products, find that some new entrant has broken a key link in the technology stack, abstracting away certain functions. These new hybrid products often come at a performance disadvantage to the discrete solution, but once they are ‘good enough’ to perform a few key tasks they can replace the incumbent solution.

I think all of this is hard to grasp from the perspective of traditional industry analysis. When the standard way of business dissolves away from some completely unexpected direction. That sounds like disruption, but it is usually just abstraction, and that is powerful enough.

One response to “Abstraction vs. Disruption

  1. whoa – was just having this very conversation, on the difference between being made obsolete (the fax being replaced by email, with some exceptions); a service being replaced by a relatively functionally equivalent service (taxi –> UberX); and having to deal with something entirely new in the context of your existing business (say, Search, or the Internet). these are all different. lately a lot of funding is going towards functionally equivalent but more convenient offerings.

Leave a Reply