Disruptive basics

This note will summarise the idea of disruptive technology.

Christensen's (1997) original formulation

Christensen studied the disk drive industry for his dissertation. Traditionally, the key performance dimension in the industry was the storage capacity of the drive. The physical dimension of the disk drive did not change as frequently. Mainframe computers required 14'' drives, minicomputers required 8'' drive, and desktop PCs required 5.25''. Christensen made two key observations. First, established companies had great trouble getting involved producing smaller disk drives. Instead new entrants typically were able to take the lead in each new size-generation of disk drives. Christensen put forward the hypothesis that the established companies listened too much to their existing customers. Mainframe computer manufacturers required 14'' drives with bigger capacity, so that is what the companies delivered. A smaller disk would have meant less capacity. The new entrants were the ones developing the smaller drives and selling them to totally new customers, in this case the minicomputer manufacturers, which started out as a niche market. Then something similar happened with 5.25'' 3.5'', 2.5'' and 1.8'' drives. Second, the new smaller disk drives improved with time and eventually became good enough for the old mainstream customers (i.e. storage capacity in the smaller sized drive became sufficiently large). At that time, the existing companies were seriously disrupted by the new entrants.

Side point. Christensen also argued that the existing accounting systems allocated all fixed costs to products. Such a practice saddled new products with too high cost allocations. This resulted in top management wrongly allocated too few resources to new products.

Christensen's work quickly became extremely influential and he wrote three books on the topic. One key problem was that he did not define disruptive technology very clearly. With each new book and with the diffusion of the original idea, people began to see disruptive technologies everywhere. Academics prefer clear definitions to be able to further research an issue. The situation was not helped when Christensen (2003) argued that it should be called disruptive innovation instead of disruptive technology.

Disruptive vs. radical innovation

It is useful to make a distinction between two concepts:

• Radicalness is a measure of how novel the technology is. This is a continuum from a very incremental (e.g. Iphone 4S, Sony Xperia Z1) to a massively radical innovation (e.g. making fire, developing agriculture). In a business context, innovations like digital photography, mobile phones, and MP3 players are generally considered radical. Since radical innovations require investment, they are normally expensive in the beginning of the life cycle. Cost decline only gradually due to economies of learning and economies of scale. 
• Disruptiveness is a measure of how value is appreciated by an niche segment and not by the company's mainstream customers. This measure does not directly deal with the technology. The disruptive technology should be inferior on a key dimension valued by the mainstream to avoid being appreciated. However it should be superior on a dimension valued by a niche segment (e.g. smaller, lighter, less powerful). 

Figure 1.

We can simplify the two continuous dimensions into a two-by-two matrix. The result is four kinds of innovation, two of which are going to be disruptive:

• Disruptive but not radical. Christensen's original disk drive example is a good example. Shrinking the components to fit a smaller frame is not really radical. Christensen also provided the example of hydraulic excavation machines replacing mechanical cable excavation machines. First their capacity was small and they were sold as add-ons to tractors to a new niche market. Many of the subsequent examples are less about technology, but more about business model (or strategy) innovation. The low cost airlines do not really use any radically new technology; instead they have a differently configured value chain.
• Radical but not disruptive. Innovations that focus on giving better performance to existing customers are not disruptive even if they are radical. The DVD player is superior to the VHS player and the jet engine superior to the propeller engine. There is no new niche market involved. It should be relatively easy for existing companies to develop products that perform better on dimensions that the existing customers already value.
• Disruptive and radical. Some innovations involve both new technology and new niche markets. Digital cameras neither appealed to professional nor amateur photographers in the beginning. Man-made polyester fabric was not very pleasant to wear in the beginning, but was more durable than cotton. MP3 players do not have the same sound quality and the files could not be bought legally. All these products initially found a niche market.
• Neither disruptive nor radical. Most activity happening in an R&D department is of the type. Over time, the result of numerous incremental innovations could be a massive change in performance. However, due to the gradual performance improvement, such investments are not disruptive when they are targeted at the existing mainstream customers. 

A few more examples:

• The mechanical cable excavation machines were initially driven by a steam engine. Once the petrol engine had been developed the manufacturers just replaced the steam engine with a petrol engine. The petrol engine was vastly superior to steam and no further modifications had to be made.
• Dell's original strategy was to sell PCs by telephone at at time when the competitors sold their PCs through retailers. This was a disruptive, but not radical. Suddenly, small investors could get the same commission rates as the big investors. When the Internet came along, it was not disruptive for Dell since it was just another way to reduce its cost.

Why do the established companies allow themselves to be disrupted?

Christensen identified a phenomenon. Many researchers have tried to explain it. Reading the research makes me think about the three blind men touching an elephant. The first describes a big flabby ear, the second describes a soft tube, and the third describes a thick leathery surface. Researchers have a similar tendency to focus on certain aspects of a phenomenon when trying to understand it. The consequence is overlapping explanations. (The explanations might be overlapping or contradictory. It is difficult to know if it still is an elephant if a fourth describes a metal like surface.) Here are a few explanations:

• Culture. Established companies that have a culture of being visionary (i.e. future market focus) and willing to cannibalise existing sales (e.g. internal competition exists, product champions exist) are less likely to be disrupted and more likely to introduce radical innovations (i.e. Chandy & Tellis, 1998).
• Resource allocation. Established companies have different R&D activities each focusing on the different components of a product. Thus they are able to improve each component. However, they often fail to innovate by combining the components in novel ways (i.e. architectural innovation). The established companies are generally bad at communication between different units (e.g. Bower 0, Burgelman 0)
• Communication patters. Exactly the same story can be analysed in terms of communication patterns instead of resource flows (e.g. Henderson & Clark, 1990).
• Organisational capabilities. Once the capabilities are created over maybe ten years, they are a source of inertia and rigidity (e.g. Sull 0, Leonard-Barton 0).

Some of the established companies did spend some resources on the new technology. Kodak's R&D department developed digital camera technology in the 1970s. Nokia's (and Sony's) R&D departments developed smartphone technology (e.g. touch screen) in the early 2000s. We can hypothesise what happened using the four types of explanations mentioned above.

• Culture. Maybe the R&D department had vision, but not top management. Or maybe everyone had the vision, but top management was not willing to cannibalise existing sales.
• Resource allocation. Maybe top management allocated some resources to radical innovation, but not sufficient amount.
• Communication patterns. Hardware and software engineers need to talk to each other to develop joint products like Apple's touch interface. Maybe hardware and software engineers at Sony and Nokia were not communicating efficiently. In particular, not enough attention was given to the software engineers because hardware had more status internally.
• Organisational capabilities. Nokia had developed a low-cost manufacturing capability. This created a stable flow of profit. Everything in the organisation becomes optimised to suit the capability.

British Airways started Go, a low-cost airline in 1998. It was started by CEO Bob Ayling. His successor decided to sell Go for GBP 100M in June 2001. Two years later Go was sold to Easyjet for GBP 374M. A totally lousy deal for British Airways. In this case the culture changed abruptly at the top and with that also the resource allocation.

Christensen effect

Andy Grove, a former CEO of Intel, described disruptive technologies from another perspective. He dubbed it the Christensen Effect. Instead of seeing certain kinds of innovation as disruptive or sustaining, we should focus on the tendency of certain innovations to disrupt. Technologies that initially perform worse and appeal to a niche market can be lethal for some incumbent companies. Calling it the Christensen Effect implies that companies can do something about their situation. They can overcome the Christensen Effect, and a few established firm actually manage to do so. In Christensen's original research there always were a few established firms that were able to do the transition successfully.

Solutions

It is easy in retrospect to determine if an innovation ended up being disruptive or not. It is much harder to do the same evaluation in real time. This had led some to argue that the established firms should not even bother trying. I think that is far to pessimistic. From the perspective of society we could say that it does not matter who is innovating as long as innovation happens. However, from the perspective of the established firms owners it certainly matters who is doing the innovation; established firms or new entrants.

The most commonly proposed solution is organisational. Christensen concluded that disruptive innovations must be developed in a new organisational unit to have any chance of surviving. Without the protection of a separate organisational unit, the influence from the existing operations would suffocate the new initiative. There are different ways of creating separation, e.g. a corporate venture capital unit, a separate business unit, a separate project directly under the CEO. Setting up a separate organisational unit would increase the chances of success, but could naturally not guarantee success.

In addition, culture, resource allocation, and communication patterns must be configured correctly. There has not been much research in this area. Maybe that is why established companies keep getting disrupted.