Will 2008 be the year of the “web computer” ?
Back in the 1990’s there was this idea of “network computers” - computers that did nothing without their connection to the network. The idea fizzled. Now we are on the verge of the “web computer”. Same idea but with better timing.
The emergence of Amazon S2 & AWS , Force.com, Facebook, and Google Apps & Gmail demonstrate the viability of software applications - corporate and personal - being served via the world wide web. Before I continue, I acknowledge the evolution of software from local installations to web based services assumed high speed internet and that is not a ubiquitous fact today.
These technologies have replaced personal computer software with web versions and provide the large data center infrastructure for building the complex custom applications traditionally relegated to corporate mainframes.
So, what does this mean for the personal computer and for the corporate software vendors ? I think there are currently two answers and they divide along the lines eluded to above - “corporate applications” and “consumer software”. Let’s look at these in reverse order.
Consumer software will evolve to take advantage of web applications. While Gmail, Microsoft Live, and Google Apps can function completely from a web browser, consumers have pushed for integration rather than replacement of their existing applications. The most anticipated features of 2007 were IMAP support for GMail and mobile sync for Calendars. The web solutions for spreadsheets and word processors were integrated with Microsoft Office and OpenOffice.org. The evolution of desktop applications as “front ends” to web applications is likely to be with us for a number of years with the separation of the two becoming harder and harder to distinguish.
I expect the PC software evolution to take on a “mergers and acquisitions” model for the next few years where the lines are drawn along four primary lines - communicating with (1) words, (2) numbers, (3) images, and music/video. “Mashups” will provide for the the areas between these. An example of what I mean can bee seen in the “words” category. There is little difference between the high priority features of a word processor, email client, and web page design tool. There is little reason for them to be separate solutions, rather, they just have separate “inputs” and “outputs”.
Corporate applications are another story. The human resource applications, payroll, compliance, and sales applications have long been heavily biased toward the server with very light - end users would say too lite - client experiences. This is a good characteristic for evolving to web solutions. The challenge to these systems has been dealing with growth, administration, and the fact they are “necessary evils” of business rather than the purpose of it. Every company - whether it be building and selling cars or books or ideas - needs the basic tools for managing the business. The only exception is the business that is “in business” to provide HR tools, or Sale trackign tools, or inventory and supply line tools. These companies are becoming the suppliers of the web solutions or they are partnering with those who are already web solution suppliers. Rather than let the business management necessities be a drag on the corporation, pushing them out to the web makes good sense.
In the “dot com” era, it is easy for startups to use web solutions because they had no legacy data or processes to contend with. They wanted to hit the ground running in their respective “hot markets” and not be bogged down by infrastructure. The old iron horses were not as nimble. But eventually all tools wear out and need to be replaced or upgraded and when they do, they are prime candidates for moving to the web. The “iron horses” are learning where and when to change.
This post is in some ways my “prediction” for 2008. I look forward to seeing how much of it comes true and how quickly. In thinking through some of the players in the “web applications” space I hit upon an interesting question of a statistical nature.
Which company uses more computer processing power every single second of each day on behalf of it’s customers - Microsoft, Amazon, eBay, Google, IBM, Apply, Sun, Facebook, or someone else ? For extra credit, what is the ordering - from most MIPS to least MIPS -for this group ?



