In the finger-pointing about how U.S. authorities might have gathered enough information to head off the Sept. 11 terrorist attacks, a shocking lack of communication came to light. A few months ahead of the attacks, an FBI agent in Minnesota raised an alarm about Zacarias Moussaoui, thinking his interest in learning how to fly large jets was related to terrorism. At the same time, an FBI agent in Phoenix had noticed that a group of Arabs had enrolled in flight school and thought he could tie the trend to Osama bin Laden. Yet the two agents never knew about each other's concerns until too late. Without corroborating evidence, both agents' concerns never went anywhere.

This sort of breakdown in sharing knowledge-which can result from what the CIA calls TMI, or Too Much Information-is so common in the corporate world that businesses will try to solve the problem by investing $12.7 billion in knowledge-management systems in 2005, up from $2.3 billion in 2000, according to research firm IDC (www.idc.com).

Yet, even as they spend all this money, many executives have the sense that they've been there and done that-with little to show for all the effort they've put into knowledge management. Is all that investment being wasted? Or is there some way to do a better job and reap the benefits of the effort?

If you look at how companies approach knowledge management, you can see that the problem is in the execution.

Companies commonly make catastrophic mistakes by falling for one of these seven myths:

KNOWLEDGE MANAGEMENT IS ABOUT KNOWLEDGE.

"Knowledge" is one of those words that sound great. Who isn't for more and better knowledge? It's almost patriotic. But the real question is: knowledge to what end? Companies that deploy knowledge-management systems hoping they will eventually stumble across a purpose for their knowledge-and there are many such companies-may be in for a long wait. The systems must start and end as all business initiatives should, with a focus on delivering top-line growth, improving operations, and increasing profit margins.

KNOWLEDGE MANAGEMENT IS ABOUT THE TECHNOLOGY.

Many companies become so focused on building the knowledge-management system that deploying the technology is all they do. And they fail. One large firm I know built the Rolls Royce of knowledge platforms, a true technological tour de force. But everyone was so busy overengineering the system that they gave too little consideration to how it would operate, to what problem the system was supposed to address, and to how it would integrate with the overarching technology strategy. In the end, the system could not keep up with the rapidly changing business, there was little flexibility to adapt to individual users, and every upgrade became a Herculean task.

THE SYSTEM SHOULD BE SO ALL-ENCOMPASSING THAT IT CAN CURE CANCER AND END WORLD HUNGER.

In fact, if you set enormous expectations, you're almost guaranteed to fail to live up to them and may be dismissed as a failure. Instead, realize that you don't have to solve every information problem in your business on the first day. You should start small, so you can demonstrate successes and develop evangelists for your efforts.

A telecom company I was involved with took the right approach. Pursuing a huge contract at a large banking customer, the telecom company leveraged knowledge-management tools to integrate the efforts of a global sales team. The team was never caught off guard through the long, arduous sales cycle and had the agility to win the deal. Based on the visibility of that success, and on the lessons learned during that initial foray into knowledge management, this telecom company has been able to extend the capability to all of its most significant sales activities.

One caveat: You need to think big even as you start small. That's because you need to make sure that the knowledge-management architecture that you begin with will still work as you expand to include other parts of your business.

THE GOAL IS TO CREATE A DOCUMENT REPOSITORY.

Certainly, document management can be a priority if employees often have trouble finding critical information or carry out redundant efforts to develop the same information. However, you must focus as much on the value and reliability of the information as on how the information is stored.

The research department of a global firm I know found this out the hard way. It put all its white papers and research reports online but found that few people used them. Then the firm built a way for people to query each other-people no longer just looked up information but could find the scientists who generated the information and ask a precise question. Employees were delighted. They made better decisions, and in less time.

YOU CAN BUY A READY-MADE SYSTEM.

Wishful thinking. Knowledge-management systems are as individualized as the businesses that use them. While there are plenty of good tools available commercially, the real issue is how those are all tied together company-wide and how they are integrated into your growth, operations, and technology strategies. If your knowledge-management program asks employees to use four search engines, three document-management systems, and six types of collaboration tools, on multiple types of computer systems, you're dead. Individual parts of your business might argue that they should be able to tackle knowledge management on their own, but you're almost always better off being consistent throughout your business.

KNOWLEDGE MANAGEMENT IS ABOUT KNOWLEDGE CONTROL.

Companies worry excessively that people will put content in the wrong place or that they can't be trusted with so much information. They add layer upon layer of approvals for contributing information or accessing it-and sap the potential of their systems.

One executive recently told me that his company had instituted an elaborate system for tracking the relationships it has with customers-then made sure that all its salesmen had access to very little of that information, for fear that they might defect to a competitor and take information with them.

In the end, knowledge management isn't about maintaining a pristine database. It's about fostering an environment in which people can ask questions like, "Does anybody know...?" Or, "Who can help me...?" This means an open system that encourages building relationships through communities and creating opportunities for personal interaction, across cubicles and across oceans.

IF YOU BUILD IT, THEY WILL USE IT.