Opinion: Real Intelligence, or Paralysis by Analysis?
- By Joe McKendrick
- November 05, 2003
Business intelligence and analytics used to be a closed process in which a small room of analysts loaded sales figures into proprietary software or onto spreadsheets to generate reports for corporate decision makers. Now, three trends are sweeping away this process, but threaten to make the IT director’s job untenable.
First, we’re moving a lot of BI/analytic processing power away from the analysts and to employees throughout the enterprise. The growing role of Microsoft in OLAP, data mining, and data warehouses on the back end is bringing down the cost of BI, and making it a far more “democratic” exercise. Anyone with the right tools at their desktop can now access and build reports that give them a picture of how some aspect of their business is performing.
Second, we’re bypassing the human equation altogether and programming decision-making into machines. There are now many instances in which performance data is sent right to systems, usually in real time, and machines either alert the appropriate decision makers or automatically make course corrections themselves.
Third, but not least, Microsoft's Analysis Services is helping many companies dive into BI/analytics for the first time. Since the technology is bundled with SQL Server, the start-up costs will be minimal.
All of these trends are good things in themselves, of course. However, we’re now turning a fire hose of information loose on our organizations, with no rhyme or reason as to how to act on it appropriately. Plus, the flow will be inconsistent – many corporate cultures still resist the sharing of information across departmental or business unit lines.
Organizations with poor customer skills thought CRM systems would magically compensate for that lack of skills. By spreading BI knowledge to every nook and cranny of the enterprise, can we also help organizations sort out their priorities, open up databases, and make intelligent decisions? That’s a tall order, even for the best of the best. Consider the experience just a couple of years ago with one of the leading high-tech giants – Cisco Systems. When the economy began to sag in late 2000, Cisco forged ahead with a frenzied production schedule, only to find itself stuck with about two billion dollars’ worth of excess inventory. Cisco executives quoted anonymously in The Wall Street Journal reverted to the classic excuse: it was a computer error. But apparently, the networking giant was relying on sales data, and not spotting double-bookings made by customers to compensate for potential delays in their orders.
An even more extensive BI/analytic system could have helped Cisco, of course. But the company probably had the most sophisticated BI system on the planet at the time, which spells trouble for the rest of us. BI/analytics pays off best when it looks at the organization as a whole, rather than in individual pieces. For years, and to this day, BI has been spottily implemented – a report here, a report there. Some of the best implementations I have come across are tied to customer call centers or sales functions, but not touching other parts of the organization. Some reporting tools, such as Hyperion, provided perspective on corporate financial data. Data mining tools, such as IBM DB2 Intelligent Miner, were run against customer and transaction data to discern purchasing patterns. Web-based tools tracked Website traffic statistics.
Vendors are attempting to bundle all these overlapping capabilities together into big, impressive packages. They are pitching BI technology as something that can be used to manage everything and anything within the organization – financial reports, customer data, inventory, production, human resources, supply chains, distribution, and physical assets. These new “total business performance management” systems take data coming in from all these areas and aggregate it into a real-time, events-based dashboard. However, demands for connecting all these pieces from across the enterprise will tie already overburdened IT operations in knots. Plus, IT cannot be asked to take on the role of cheerleaders for cultural change when business unit managers resist donating their data to the cause. The BI tools coming on the market are getting more powerful, less expensive, and do great things. But, for the foreseeable future, we still need to keep those boots on the ground.
Joe McKendrick is an independent consultant and author specializing in surveys, technology research and white papers. He's a contributing writer for ENTmag.com.