« Job Titles and Business Cards | Main | Ubiquitous Cell Phone Coverage »

July 19, 2006

Misunderstanding "The Innovator's Dilemma"

Mini-Microsoft, at the end of this article, describes Clayton Christensen's book The Innovator's Dilemma as the book that "tells you how it's sometimes a really bad idea to do what your customer wants."

For some reason the point of this book is often misrepresented. So, since I like the sound of my own typing, here is a list of things that the book does NOT say:

  • It's best to ignore your customers.
  • Cool, newfangled companies and technologies always beat old, established ones.
  • Cheaper products win out over expensive ones.
  • Linux is going to destroy Windows.

OK, the last one is somewhat facetious; I'm using it as an excuse to link to an article I wrote (more than 5 years ago? Eek) about how Linux is not a disruptive technology to Windows. This does not necessarily mean that Linux isn't a threat to Windows, just that it doesn't fit the classical disruptive pattern that Christensen outlines.

So what is that pattern? Well, it includes the following steps:

  1. A market has a leader offering a successful product.
  2. The market leader asks its customers for suggestions on how to improve the product.
  3. The company responds to the suggestions, which lead it to make the product more feature-laden and expensive (a direction Christensen calls "upmarket").
  4. Meanwhile, another product comes along which is "downmarket" of the existing product--enough so that initially it isn't seen as a competitor to it. The new product does attract customers but is considered unappealing by customers of the existing product.
  5. But, over time the new product improves until it is "good enough" to satisfy the needs of mainstream users. The existing product can stick around in a high-end niche, but the new product has taken over its main market--"disrupting" the old product (and the retirement plans of the old company's executives).

That's the way it works. The key is the customer-driven blind spot: the company is doing the "right thing" (listening to its customers), and while this works in the short term it fails in the long term as the customers lead them astray.

So, let's consider the example that Mini was discussing. This was the decision of to remove the folder password-protection feature from Windows, soon after it was announced. Is this a case where a company should ignore its customers in order to avoid a disruptive technology? Clearly not: it has nothing to do with that. A disruptive scenario here would be that Microsoft added increasingly complex folder management options, and then someone came in with a simple add-in that only did the one thing that people wanted, and this caused something bad to happen, except it's such a minor thing, and the add-in would be free, etc...anyway like I was saying, it doesn't fit the mold for disruption.

Keep in mind that the disruptive technology pattern is just one way that new technologies can displace old technologies. Consider the case of Google knocking off its search competitors a few years ago (at least five, since I mention using Google in the article I linked to above). Google didn't disrupt AltaVista et al by moving upmarket with a simple product that became "good enough"; they just marched in with better technology. This doesn't mean that the displacement didn't happen (since it obviously did), just that it was done with a more standard approach: having a better mousetrap.

(You could claim that Google's (or whoever-thought-of-them-first's) use of text ads was a disruptive technology relative to banner ads, but I would say no to that also; it's true that people were thinking of bigger and more annoying kinds of banner ads while text ads were simple, but the problem for search was that nobody had figured out how to monetize the results. Text ads in search results didn't replace banner ads, they replaced nothing. Or, perhaps you could say that contextual text ads in search results replaced non-contextual text ads in search results, but again that was better technology, not a disruption by "good enough".)

As a final exercise, let's consider whether citizen media (e.g. Scoblecasts and all that stuff) are going to disrupt traditional media. I think this is an interesting question, because it's not clear that they serve the same purpose. It boils down to whether you view traditional media as by definition providing something that only it can provide. In other words, you could argue that traditional media provides access (as in, "access to events that are limited enough that only traditional media gets in"), stars (as in "stars that are big enough that they can get distribution through traditional media"), etc. If you believe that, then the long tail can't disrupt the short head because the people desire the short head precisely for its short-headedness. BUT, if you believe that traditional media is mostly just filler for people to consume between dinner and bedtime, and all people need is 15 frames a second and a through line, then it's entirely reasonable to believe that the long tail can become "good enough" to whip around and knock the short head off its perch.

Posted by AdamBa at July 19, 2006 09:50 PM

Trackback Pings

TrackBack URL for this entry:


I think you're correct that Linux isn't going to destroy windows.

Linux is rather like the Visigoths, who sacked Rome after it had collaped under its own weight.

Ten years from now, when Windows has retreated to the stock rooms of dingy tire shops where DOS 2.0 still reigns today, and people ask what brought down the Evil Empire, the answer will be that MS simply outgrew the competence of its managers, and was clearly doomed to collapse, leaving Windows 95 as the high point of its existence.

Linux will simply rush in to fill the void. It's not that it's particularly good, but it is at least securable and reliable, two things that really matter when your machine is exposed to the internet, and which Windows will simply never achieve.

Posted by: Some Guy at July 26, 2006 05:15 AM

The manager of one of the largest pension funds invested in Microsoft said Microsoft was creative, but just not very entrepreneurial. For too long R&D at Microsoft has stood for "rent & deploy." We go out to buy a technology or company and then deploy it under our brand. Whatever happened to REAL risk taking besides betting that Vista would be out on schedule?

Posted by: Wallen at July 29, 2006 03:29 AM