Business Intelligence Reboots Yet Again
Business intelligence applications, software that is supposed to make sense of data in disparate databases and then spit out the answer to every question about your business, is trying once again to redefine itself. The business intelligence (BI) industry reminds me of a baseball player who broke into the major leagues with star potential, but ends up being shipped, along with his potential, from team to team for a half-dozen years.
The potential for BI is obvious for all to see -- automatically extract, read and interpret tens of thousands of blog posts and customer complaints and figure out why customers hate your company, or simply use sales data to figure out how many yellow plus-size dresses need to be shipped to the North East. But the reason your bank still infuriates you and your favorite store still doesn't get clothes in your size isn't because they've never heard of BI. Chances are they even use some; and that, in a nutshell, is the issue with BI.
There's also the fact that every major bank in the country was using BI, but still churned out loans to unqualified buyers, because they built logic into their predictive models that assumed that real estate prices would grow by six percent per year... forever. That's not a software problem, it's a people problem. Or as IDC's search and litigation discovery guru Sue Feldman put it to me this morning at her research firm's BI conference in New York, "you can't assume the future is like the past."
The BI industry has been through several iterations of itself, from the early days when business analysts had to ask someone in IT to create reports using complex queries that took so long to build, the reports were anachronistic by the time they landed on the analyst's desk, to a period when reports were somewhat easier for IT people to create, and something called OLAP cubes were created, giving business analysts a modicum of flexibility in how they read the data, to today, where increased computing power and better tools for business analysts allow almost real-time ability to crunch and analyze data. But the reports are still difficult to use as decision-making tools.
Some BI vendors like Attivio, Qlikview, and Spotfire (which is owned by financial services software vendor Tibco), are trying something slightly different, which is to allow business analysts to use familiar graphics like line graphs and pie charts as the starting point for making queries, and returning results in reports formatted like those charts as well. I also met with a new BI vendor last week, Mellmo, which has entirely reworked how data is organized on the screen, probably revisiting the whole idea of how reports should be read for the first time since reporting transitioned from mainframe computer print-outs to the PC. As Mellmo CEO Santiago Becerra noted, that's a long time between paradigm shifts. Mellmo's application is built for use on the iPhone -- and reading reports on a tiny screen is a great incentive for rethinking how users read reports -- but Becerra thinks the technology will also be adapted for larger screens.
It looks to me like data visualization is the next big push for BI vendors, and that should certainly help business analysts make better decisions. And the hunger for this kind of technology is palpable -- according to Feldman, BI search and discovery software sales are growing at a twenty five percent clip this year, despite -- or perhaps thanks to -- the recession. Henry Morris, senior vice president of software research at IDC, said during his keynote presentation that "not knowing what you know wrecks havoc with an organization in a way it might not in more prosperous times."
Trouble is, BI may have helped speed the recession along the way, so it's hard to know whether to root for these guys to be more successful or not.
[Image source: Wikimedia commons]