Conference Kernels: BI Mega-Trends for 2006
Using the TDWI Business Intelligence Framework, research and services senior manager Philip Russom outlined four ‘mega-trends’ to watch in 2006.
- By Eric Kavanagh
Whenever a new year approaches, pundits and professionals alike wonder aloud what to expect from the coming 12 months. In laying out a roadmap for what 2006 might bring, TDWI’s senior manager of research and services, Philip Russom, used his Thursday keynote at TDWI’s World Conference in Orlando this month, to outline four "mega-trends" to watch.
To put these trends in context, Russom referenced TDWI’s business intelligence framework, an overarching matrix that details the component parts of a BI environment, and shows relationships between the various elements. "Four issues hitting most pieces of the BI framework are: size, speed, interoperability and economics," he said.
On the first mega-trend, Russom noted that size applies to a host of issues, including: "the size of data sets you’re integrating through ETL or some other process; could be the size of storage, the amount of data in a warehouse, data mart, ODS; but not just the size of data: also the size of your user base; report consumers; report producers; and the number of reports. So, scalability is an issue associated with size."
For mega-trend number two, Russom noted that speed deals with more than response time. "Frequency is an issue," he said. "Companies are fetching data and integrating into a warehouse on a more frequent basis. We must get beyond batch processes; that’s where speed affects data integration. And if you haven’t yet addressed some form of real-time data warehousing, you will."
The third mega-trend, interoperability, is largely "ignored," noted Russom. "There will be cases where different pieces of the framework must work together. Yes, ETL must work with EAI, with your reporting platform, your scorecard and dashboard reports. In some cases, the only way to get at application data is through EAI."
On the topic of integration, Russom further elaborated: "Today’s best practice is to use multiple integration technologies to feed a data warehouse, being sure that they don’t overwrite data that they all touch. The future best practice is to hand data directly from one integration technology to the next, which is fundamental to service orientation and real-time BI.”
As for the ever-important issue of economics, Russom pointed to trends of product pricing versus data explosion. He noted that price points on hardware and software are going down, while the proliferation of data stores—and resulting issues related to data quality—require additional investment for consolidation and cleansing. He said while some scrimping can be healthy, organizations that take economics to the extreme will compromise their ability to harness the value of business intelligence.
After teasing the audience with a brief showing of the Holy Grail—namely, optimization of speed versus size—Russom concluded his keynote with a significant caveat. "IT containment is here to stay," he said. "Even if the global economy picks up, expect to rationalize all IT investments economically—including BI."
Eric Kavanagh is the president of Mobius Media, a strategic communications consultancy. You can contact the author at firstname.lastname@example.org.