How often have we heard about the growing importance of using business intelligence to unlock the insights bound to be contained in corporate data? It's a perennially popular theme, one that has kept BI near the top of CIO priority lists for several years running.
Why then is BI treated as "just an application" as Dave Linthicum contends on eBizQ? BI requires some heavy-duty architectural planning to be truly effective, he writes:
You need to consider how the information in the operational data stores will be aggregated, replicated, or abstracted, or all of the above. Thus, you need to consider data integration patterns, including latency and logging. Also, you need to consider data governance which should be dealt with at a systemic level within the architecture.
Linthicum's larger point is that all important business systems, including BI, should be incorporated into enterprise architecture planning. Yet too often they are not, resulting in awkward data integration workarounds and systems that don't live up to their potential.
What's needed is a more holistic view of how organizations use data now and how they want to use it in the future (both near and longer term). IT Business Edge blogger Loraine Lawson wrote about this just last month in a post titled . Among other knocks, a platform approach makes it tougher to attain ROI, say some experts.