Thursday, September 08, 2005

Web OS: The Battle for Dominance

My prior post addressed the brewing battle of the two dominant computing platforms: the Web OS vs. Desktop OS. The value of the platform is the ability to aggregate developers/applications and end-users and then make these two groups use your platform (traditionally the operating system) to access the each other. For example, if you’re an end-user and you want the largest selection of applications, you run Windows. If you are a developer, and you want access to the largest collection of end-users, you write your application to Windows.

This dynamic then feeds upon itself according to the Law of Increasing Returns (AKA the Virtuous Cycle). More developers = more applications = more users = more developers…

Traditionally, the operating system has been the chokepoint between developers and end-users. By owning this chokepoint, IBM in mainframes and Microsoft in PCs, have been able to extract tremendous value and leverage. Both companies exploited this leverage to dominate their respective markets.

We have also seen a number of competitors attempt to compete with Microsoft as platforms. Novell, Sun/Java and Netscape have attempted to define a platform alternative. All three have attempted to woo developers and end-users in an effort to assemble their own platform. All three have failed for one reason or another, not the least of which was Microsoft’s ruthless competition.

Let’s look at what it takes to assemble a platform:

1. End-users.

If you are going to knockout Microsoft, you must have either more users, or a trajectory that, if maintained, provides the promise of more users. At the very least, you need to provide access to a roughly equivalent number of users.

2. A Superior Solution for Developers

This is comprised of a number of things.

  1. Tools: You must provide excellent tools, so that developers can work in an environment they like, and one in which they can assemble cutting-edge applications
  2. Opportunity to Pioneer: You must enable developers to build unique and interesting applications, something they couldn’t build on the competing platform.
  3. Opportunity to Profit: Developers—with the exception of open source—want to make money. If the costs to build, test, package, promote, sell and support their product is too high (as with mainframe software) they will look for a less costly business opportunity.

[Note: Clearly, the operating system isn’t the only platform. A good example of a non-OS platform is Oracle’s database. It followed the formula described above, by building both an end-user base and a developer base. This then started the virtuous cycle that has fueled the company’s growth.]

The big question is whether the web now provides the characteristics necessary to become the dominant platform. The answer is yes. The web has the user base. The web provides developers with the opportunity to pioneer, especially in the areas of community, social networking, collaboration, etc. There is also the opportunity to profit on the web because your web application has access to a tremendous web user-base, while virtually eliminating traditional costs associated with publishing, distribution, marketing and support. Any kid with a computer and a broadband connection can now launch a web application.

The ability to both pioneer and profit are demonstrated by web start-ups like Yahoo, eBay and Google that have become multi-billion dollar companies.

The current web application development tools are server-centric. And building server-centric applications on the web has been sufficient, until now. Now the game is shifting. Now applications, in an effort to battle head-to-head with desktop applications, must be more graphically appealing, more responsive and more powerful. In short the user experience needs to mimic the user experience of desktop applications.

For this reason, web development tools (languages, IDEs, frameworks) need to: (a) provide more desktop-like functionality (e.g. drag-and-drop); (b) facilitate client-side processing. This is happening. In some respects, it appears to be a race. Microsoft is racing to leverage web resources, while the web developers are racing to leverage client-side resources.

Handicapping this race is interesting. Microsoft has the edge in user-base, but the web is closing fast. This is a new and unique challenge for Microsoft. Microsoft’s tools and developer programs still lead, but there are some very good tools emerging on the web side. Where the web has the edge is in the opportunity to pioneer and the opportunity to profit.

People are building very cool web applications that you could never envision on the desktop. Applications like Flikr,, Facebook, various mashups, this stuff is really new, unique and interesting. The killer advantage of the web is its low cost structure and the ability to build something cool and watch it spread, simply by word-of-mouth. This results in a tremendous opportunity to profit.

Web development tools are evolving. They are beginning to enable developers to leverage client-side processing more easily. This will result in even better and cooler applications. As a result, the web will evolve as a platform that more directly competes with Windows.

The question is whether one company will emerge as the owner of this platform. By virtue of the web’s distributed nature, it looks like we won’t have a single chokepoint on the web platform. Instead we’ll have alpha-dog companies, like Google, who will lead and influence the evolution of the platform, but there won’t be a single owner of the platform.