Keys to intelligence success

Lately I've been glued to C-SPAN, watching the Sept. 11 commission's hearings, and I found the testimony given on April 14 riveting. Witnesses from both the FBI and the new Terrorist Threat Information Center (TTIC) echoed the complaints I've heard time and time again from corporate IT and business management.

Executives and analysts don't want disconnected, formulaic reports from each department within each unit. Instead, they want summaries that consider facts from all sources -- even facts they aren't trained to comprehend – that they can distill into high-level views supporting quick and smart decisions. That's the definition of intelligence. Anything less is just data.

Granted, the FBI isn't gathering the same kind of data as your IT shop. But the FBI's knowledge customers, which include the president, are tasked with making split-second decisions that rely on the accuracy, timeliness, and relevance of the information given to them.

The rising interest in BI technology is evidence that corporate and organizational IT is facing demands and criticisms markedly similar to those confronted by the FBI. The volume of data that feeds business decisions will rise by orders of magnitude in the coming years, and yet decisions will have to be made more and more quickly.

To those outside the process, business facts are the same as business intelligence: "Can't you just tweak the parameters of a few reports?" But buying that myth -- and it's sold aggressively by less reputable vendors -- will result in a torrent of lousy business decisions.

Instead of handling intelligence as a purely technological problem, a business intelligence strategy starts with deep organizational and cultural changes that pry data loose from those who collect and keep it. It calls for additional staff dedicated to intelligence-gathering and analysis. Leaders of business units who will be developing their own intelligence need to link arms to ensure that both facts and intelligence are shared.

In the FBI's case, the bureau is moving its counter-terrorism operation into a building with other agencies involved in fighting terrorism, including the CIA. Traditionally, these agencies would each protect their turf by asserting exclusive jurisdiction and hoarding data. That stovepipe culture is a problem familiar to too many businesses, and it must be dismantled before any BI strategy can be effective.

One crucial BI stage that must be driven by IT is the creation of a single mega-repository to collect all data that might ever be used for intelligence analysis. This all-knowing warehouse (or whatever model serves you best) should not allow data gatherers to opt out; policies should determine which data will be published by each unit. Only information of the most sensitive nature (for example, employee salaries and customer credit card numbers) should be withheld.

To put this in perspective, in Washington's case the mega-repository is managed by the TTIC, whose systems can reach directly into the FBI's databases. That may seem to put a lot of power in one place; but at the same time, the mega-repository instantly becomes a searchable portal for all the units that contribute to it.

The mega-repository doesn't exist to analyze data or make recommendations. A cohesive intelligence strategy cannot deprive units of their power to make decisions, and even to disagree with each other. Such disagreements are crucial elements of findings presented to senior staff. BI’s aim is not to eliminate human analysis. Rather, it recognizes that we have most of the skill and technology we need to excel at information gathering. It’s time to move to the next stage.