The Undoing of Disruption

A strange article. Many of the things cited as proof that the theory has poor predictive power are really dependent on how narrow of a scope you're willing to ascribe to industries and sectors.

For example Christensen stated that the iPhone wouldn't be a disruptive technology because it is essentially a very high end phone. This turned out to be a predictive miss on Christensen's part, not because the iPhone succeeded in spite of being a high-end phone, but because Christensen mis-categorized it: it's a very low-end personal computer.

Furthermore, the article uses the HDD market as an example. It states that 1.8" is not the dominant form-factor today as Christensen predicted it would be (instead 2.5" and 3.5" are). However, the article doesn't state whether the analysis took into account the SSD market (I don't have data on the SSD form-factors but all of my laptops have a 1.8" SSD, for what it's worth).

The main problem with the theory's predictive power is that it hinges on being able to accurately categorize companies/products as being part of a particular _industry_ or _sector_. Where does the cell phone industry end and the personal computing industry begin? Are HDDs and SSDs considered to be in _different_ industries, and why? What about railroads and air travel? The lines between industries are blurry at best.

Obviously everything is connected, and the theory is more about predicting the long-term macro-economic interactions between companies/players and the market. Instead of drawing S-curves and looking at industries and sectors (which are really just arbitrary, imperfect categorizations that don't take overlap into account) one should look at which players are stealing other players' low-end customers. That's likely to be indicative of a disruption-in-progress.