Commoditizing On-Demand

It is inevitable. In economics, commoditization is a routine part of the life cycle, the opposite end of the spectrum from innovation and as natural as life and death. Sometimes commoditization results in the elimination of a product or idea, but at other times it merely means subsuming the innovation into something larger. I believe we are witnessing the latter in the case of on-demand technology.

Let me be clear, I said “technology,” not the movement to lighter, easier-to-use and less costly software or the business processes it spawns. But it is hard not to notice how many vendors are now taking up on-demand lite — adopting aspects of the technology to make conventional applications perform better and appear more acceptable to customers.

I don’t know if anyone writes an application in traditional client server anymore. Standalone applications like the one I am using to write this column were built to run in a window, but there are alternatives that live on the Web and run in a browser, so the trend is clear. Many more new applications of all sorts are being delivered in browsers, and there is the wedge into on-demand.

Data Migration

Once an application operates satisfactorily in a browser, some of the heat is off the developer and the IT group. The user might be unaware of where the back end of the application is, and that’s as it should be. The back end could be on the Web or it could be down the hall in the data center. The adoption of such on-demand innovations as browser interfaces and stateless computing make it easy for a more or less conventional solution to look, well, avant-garde.

As a result, we see vendors trying to blur the lines saying that their products run everywhere — on-demand, in a third party data center or your own — it’s all good. And perhaps it is. But an application that runs in a private data center may still be rather hierarchical instead of offering a more networked topology. The difference is significant.

If Twitter ran on a bunch of private servers instead of running in the cloud, you might have great insight into the lives of the people down the hall (and maybe too much), but keeping up with your friends in New York, Chicago, Austin and San Jose might be a bit more clumsy. You might need to wait for the batch update before knowing something, and then what’s the point?

Back End Matters

We’re not just a global economy these days; we’re an instant-on, always communicating global tribe. In myriad, inexplicable but very real ways it matters what we think and what others think about what we think — ad nauseam. Each thought is a pebble dropped in the pool, and we only know that the ripples go out but not always whom they affect.

When on-demand meant delivering an application as a service over the Internet and nothing else, you could make a point that where the back end lived was not important and that multi-tenant architecture was not very important, but that argument is less persuasive now. It was not totally persuasive back then because multi-tenant offers so much savings over other forms of computing (even hosting), and it is not persuasive now because it ignores the importance of leveraging the cloud for staying in touch — with friends as well as customers and business partners; in short, social applications.

A few years ago I came up with a category that was uniformly ignored by the market — WebNecessary applications. These applications were supposed to leverage the Internet for much more than simple communication among and between servers. I figured the category would fill itself up with specialized applications for various business processes. I did not foresee that social networking applications would be the catalyst and that they would adhere to everything like gravy on a turkey. So much the better. Social networking is making our routine business applications WebNecessary.

With the arrival of cloud computing, we’ve hit a new marker in the evolution of computing. Some elements of on-demand technology are being subsumed into conventional computing, and this will serve as a bridge or paradigm extender for older approaches. It will give time for the laggards to change rules and regulations that may be preventing some organizations from leveraging the new technology. But make no mistake about it, cloud computing is rapidly becoming the new standard.