“Traditional business approaches as we know them have not proven really successful,’’ said Boris Evelson, a principal analyst at Forrester Research, at the association’s “Data Foundations and Business Intelligence” seminar at the Westin Hotel in Times Square.
The problem: All so-called “business intelligence” systems are based on a complex setup of computing methodlogies, processes, architectures and components, from dashboards to portals to analytics, that are based on one fundamental structure: a database.
This in turn means that every question that might possibly require an answer be figured out in advance, when setting up the data model, Evelson said. Which means “pre-discovery” of questions.
Any new question that emerges in “post-discovery” will need adjusting many parts of the system. A simple change in assumptions in a Value at Risk query, for instance, he said, could take days to set up a new query, make sure the needed data was in the system or could be added to the system, adjust other parts of the system and then deliver a report.
This is why systems that can do “ad hoc analysis,’’ by putting needed data in readily searchable memory are gaining ground, he said.
"By the time you get the answer, you've forgotten the question," said Venkat Mullur, senior director for industry solutions at TIBCO Spotfire, which provides software that is designed to encourage on-the-spot queries and answers.
The company’s software allows investment banks, hedge funds or other users to adjust scales in a dashboard, for instance, to change the nature of relationships between different sets of market factors.
For instance, the head of a trading desk worried about the potential impact of the “Greek contagion” last week (or this) could change the level of correlation between changes in the prices of Greek equities and U.S. equities, get an instant scatter plot of the revised correlation, and act accordingly.
“At the end of the day, analysis takes place at the speed of thought,’’ he said.
This article can also be found at SecuritiesIndustry.com.
Tom Steinert-Threlkeld is the editor-in-chief of Securities Technology Monitor.
Be the first to comment on this post using the section below.