SAST Wingees

Could Apple mess up, again? (with the iPod)

Updated again Sep 19, 2006: Nick Carr wrote an interesting post on digital music expressing some similar arguments against Christensen’s theory. He doesn’t read my blog, of course. Well, Nick read the post and was kind enough to include a link in the comments  [#22] section of his post. Please read the other comments as well. There are some good ones.

A few months ago, Clayton Christensen of “Innovator’s dilemma” fame argued in this Business week article titled “How Apple Could Mess Up, Again” ,  that Apple will lose its dominance in this market like it did in the PC market.  I have been thinking about this article since then. I am surprised that the usual defenders of the faith  (gruber, macslash, apple matters) haven’t come out swinging against this article.   So I decided to take this upon myself. As a person with a lot of passion for business strategy, I am a big fan of Christensen’s path-breaking theory. But I have come to conclude that Christensen’s theory of  interdependent product architecture versus modular product architecture doesn’t really seem to fit the computer industry well. I noticed it first in his application of this theory to the IBM PC and Microsoft’s IE browser.  I will write why that is a flaw later. Coming back to the Apple iPod, Christensen again applies his theory and says that  because the iPod is based on an interdependent architecture, it will lose to a competitor with a modular architecture. Christensen defines interdependent architecture thus in his book “Innovator’s Solution“:

An architecture is interdependent at an interface if one part cannot be created independently of the other part  – if the way one is designed and made depends on the way other is being designed and made. When there is an interface across which there are unpredictable interdependencies, then the same organization must simultaneously develop both of the components if it hopes to develop either component.  Interdependent architectures optimize performance in terms of functionality and and reliability. By definition, these architectures are proprietary because each company will develop its own interdependent design to optimize performance in a different way.

 He defines modular architecture thus:

A modular architecture specifies the fit and function of all elements so completely that it doesn’t matter who makes the components or subsystems, as long as they meet the specifications. Modular components can be  developed in independent work groups or by different companies working at arm’s length. Modular architectures optimize flexibility, but because they require tight specification, they give engineers fewer degrees of freedom in design. As a result, modular flexibility comes at the sacrifice of performance.

The basic premise of Christensen that the iPod belongs in the interdependent architecture category is wrong. It actually belongs in the modular architecture category, as we will see shortly. Apple has actually masterfully integrated 3rd party modular components to build the iPod. There have been numerous teardowns of the iPod to show what is inside. This teardown by Jefferies & Co’s analysts shows the entire Bill of Material for the iPod and there is not a single Apple proprietary component. 

You can also read this Wired Magazine story “Birth of the iPod” that talks about how PortalPlayer became the brain inside the iPod.  Its pretty easy to see that the story is somewhat similar to how the IBM PC was put  together first by using 3rd party components. 

Of course, there are some key differences. Further, consider the fact that the iPod is entirely manufactured by ODMs (Original Design Manufacturer) Asustek and Inventec – that is, Apple supplies the design and these ODMs manufacture the product.  I am sure one of these will be willing to produce a knock-off without violating Apple IP. Toshiba, suppliers of the mini hard drive for the ipod,  themselves produced a knock-off in the Gigabeat line.

In fact, the iPod is so easy to hack (not for the lay person, of course) that a completely new firmware called Rockbox has been written for it by open source enthusiasts. Some have ported Linux onto the iPod as well. But Rockbox is more specifically suited to a MP3 player operation than a general purpose operating system like Linux. Many companies including Dell  (Dell DJ) and others have tried to compete against the iPod and given up.

Actually, Dell is the king of the Modular Architecture camp using its super-efficient supply chain to drive the costs down. It tried the same game with the MP3 Player, but couldn’t  compete. Some argue that the success is due to its iTunes software. I started using the iPod before iTunes for Windows was released and I have only 50 songs bought from the iTunes Store. But I have enjoyed  and will enjoy using the iPod regardless of what PC software accompanies it, as long as it works. Apparently, the average iPod has only 25 songs from the iTunes store.  So clearly, its not that big of a deal for most iPod users.

iTunes software is definitely a complementor, but the main success of the iPod is due to its unbeatable user experience and the cachet that Apple has been able to establish around it. Reminiscent of Alfred Sloan’s GM’s “A car for every purse and purpose” strategy, Apple has clearly out-innovated the competition by adding features, new models (Mini, Nano, Shuffle, Video, Photo) and cut prices faster to expand the lead. Another interesting thing is making the iPod into a hardware platform, where 3rd parties (Bose, BMW, Griffin, Nike…) can sell add-ons but tied to Apple’s proprietary dock port (the hardware equivalent of a Microsoft Windows API type software lock-in).

Apple has also added a few more proprietary frills in terms of DRM and AAC format. But none of this really prevents another competitor from producing a better knock-off. So this whole interdependent architecture theory and the fact that a modular architecture company will come in and beat the iPod becomes more and more unlikely as time progresses. In my opinion, if someone has to beat the iPod, they have to come up with a better business model.

And that may come from  Microsoft, the shrewdest business model innovator in the technology planet. In fact, Microsoft, has already made its first attempt, by using the subscription-based model as an alternative to the iTunes per-tune model.  Somehow, that has not succeeded probably because of the 25 songs per iPod situation that I mentioned above.  People simply use the iPod to manage their legacy music collection, atleast for now. 

My hypothesis is, that Microsoft will come up with its next attempt and that will be based on how it beat the other user-experience champion Palm. Palm is the most recent casualty of the approach Christensen advocates in the Business Week article quoted above – Apple must license iTunes-inside to all MP3 player makers, which is exactly what Palm did. It will be a long one and I will cover it next week.  Stay tuned.

References:
1. Bill McCoy of Adobe wrote a while ago on why the interdependent architecture theory doesn’t fit IT in general and iPod in particular. I googled this and found it at the time of writing this post. I came up with this same thought independently of Bill.

2. iPod Juggernaut by John Gruber.

Flashback:
1. I had argued a while back that Apple should open up the iPod API.
2. My view of how Apple should attack the cellphone/smartphone market using iPod SDIO. Still the most popular article on this blog.
3. My view of how Apple should attack the Wintel marketplace after its switch to Intel microprocessors. 

 Technorati Tags: Apple iPod, Interdependent Architecture, Modular Architecture, Strategy & Business Models