Combing the Business Intelligence Hairball

I’ve been known to use the expression, "combing the IT hairball," to add vigor to my stump speech about the potential role of on-demand in streamlining IT portfolios — thereby paving the way to improved process governance.

I now find independent support for my bleak assessment of the mess to be found in too many enterprise IT environments. Research released today by Reg Developer indicates that

In a large company environment…it is normal for literally hundreds of business applications to have accumulated over the years, often with significant redundancy but little integration between them. We found in our study, however, that even smaller organisations suffer from the consequences of information fragmentation, so it’s a problem that is endemic in pretty much any organisation that has invested in automation over an extended period of time.

When you look at the factors that drive successful use of business intelligence in organizations, it’s clear that an on-demand environment is an ideal Petri dish in which to grow a BI culture. An on-demand software platform opens the door to BI upsides while lowering associated IT burdens, in three ways:

  • Rapid deployment of on-demand systems builds a buzz of end-user acceptance during early stages of a broad-based BI initiative
  • Ease of integration, whether at the level of the API or at the "mashup" level of the user interface, makes an on-demand platform an ideal breadboard for experimentation and refinement
  • Granular privilege management in on-demand applications makes it convenient to restrict the use of sensitive information, while at the same time offering timely and relevant data to every member of a team

When people talk about the reasons that business intelligence is difficult to drive across an enterprise, their arguments reflect at least two levels of legacy. You hear in their arguments the echoes of massive data silos, harking back to the first generation of enterprise IT; you see in their objections the kaleidoscope of confusion that arose from personal data collections and uncoordinated tools during a generation of stand-alone personal computing.

The future of business applications should shed this baggage. It should offer broad access to current information from diverse sources; it should deliver that access in personalized views that exploit the universality of the Web. (Why not come to Dreamforce in September and learn more?)

July 27, 2007