Are you ready to cede control of your business-critical data?
 

Information Is No Utility

Human beings — those living in the Western tradition, at any rate — have always been obsessed with endings. Perhaps something in human nature makes us perceive an “ending” as being more dramatic than its inseparable Siamese twin, the “beginning.” Winston Churchill famously turned this inclination on its head in his wartime admonition to Britain’s despondent citizens: “This is not the beginning of the end, it’s the end of the beginning.”

IT has seen more than its fair share of “endings” in the last decade: The End of the Mainframe, the End of Procedural Programming, and the End of Client/Server, just to name a few. Most of these declarations were not fulfilled. Rather, they foretold the gradual emergence of a new synthesis in which the “dead” paradigm continued to play an important but transformed role. It’s very likely that the latest and perhaps most dramatic “ending” of them all — the End of Software, as predicted by the emerging application service provider ASP industry — will take the same course.

This Year’s Model

As the ASP industry and its allies would have it, the “information utility” model applies equally to desktop productivity apps, enterprise resource planning (ERP), and customer relationship management (CRM); presumably, owning your own data center will make as much sense as owning your own water supply. The ASP partisans even foresee the demise of integrated software itself; in this scenario, customers will assemble application components on the fly and pay for them on a per-usage basis.

The information utility model, from which analysts predict a $2 billion market will arise in the next few years, has impressive admirers. Oracle now offers hosted versions of Oracle Applications through Oracle Business OnLine and plans to make its business intelligence and decision-support products available as well. PeopleSoft has a close relationship with enterprise ASP Corio, and Qwest Communications — shaping up as an enterprise hosting heavyweight — has forged agreements with SAP and Siebel Systems. Even Microsoft is riding the trend; it recently announced a pilot program that will help ASPs “net-source” apps that run on Microsoft platforms, and at the other end of the spectrum, plans to offer hosted versions of Office.

All well and good. With increasing standardization and huge increases in affordable bandwidth, smaller companies can now plausibly afford to ante into the ERP game and larger ones can reduce their IT outlays. However, if you follow the information utility model to its logical inclusion, eventually you will encounter an inescapable quandary: You will have to relinquish control of your business-critical data.

That point is precisely where the information utility model breaks down. Data isn’t fungible like electricity or water. Perhaps you’ll permit an ASP to control your transactions, but will you permit your data warehouse to live there? And how would you feel about using applications to which your competitors have equal access to do customer segmentation analysis on data that you don’t directly control? My best guess would be: not good.

The More the Merrier

You could plausibly argue that a small or even mid-sized company has no business implementing R/3 internally, but for larger companies that live on the competitive edge, keeping business intelligence, CRM, and decision support “in the family” has obvious value. Thus, we won’t see the End of Software for a long time to come. Instead, wider access to enterprise applications, albeit of the hosted variety, may broaden the appeal of business intelligence processes — the only true source of advantage on a level playing field — over all. In my opinion, that’s a happy beginning, not an ending.