Experts divide such critical know-how into two parts: explicit and tacit knowledge. The explicit kind refers to information that can be easily explained and stored in databases or manuals. Tacit knowledge is much harder to capture and pass on because it includes experience, stories, impressions and creative solutions. Tacit knowledge is also much harder to get from people because it accumulates over years of experience, and a scientist or salesperson may not even know how to verbalize it.

Dorothy Leonard, professor emerita of business administration at Harvard Business School and co-author of Deep Smarts: How to Cultivate and Transfer Enduring Business Wisdom, argues that companies and government agencies should concentrate on re-creating tacit knowledge, rather than focusing only on transferring it. For example, if an experienced scientist plans to retire in a year, the pharmaceutical company where he works should have a younger researcher shadow the scientist and work side by side. "In this way, the younger scientist will learn not just the facts, but the method of diagnostics," Leonard says. "Databases are not a complete waste of time, but it's a mistake to believe they are transferring knowledge."

In some cases, it's easy to see that the loss of a key employee, or group of employees, will affect a company's strategy and bottom line. At engine manufacturer Rolls-Royce, for example, managers - when faced with the impending retirement of a veteran systems engineer - calculated that the engineer's retirement would cost the company $US400,000 in the first year, says Colin Cadas, team leader for design technology at the UK-based company. Cadas based the calculation on the number of employees whose productivity is affected when the system is down and the average time the system is unavailable. Using that calculation, managers could then justify the knowledge acquisition activity before the engineer (the primary troubleshooter for that system) left. The process also guaranteed increased training for younger engineers before the retiring engineer left the company. "For every knowledge retention project we do, we have those involved work out the business value to the organization," Cadas says.

Stop the Bleeding

Rolls-Royce faced a crucial test in April 2003, when British Airways and Air France ended service of the Concorde supersonic jet, citing diminishing passenger numbers and rising maintenance costs. Rolls-Royce, which had maintained the supersonic Olympus engines since the planes started jetting rock stars and business titans across the Atlantic in the 1970s, realized it needed to act quickly. And managers knew this specialized knowledge was crucial to securing future opportunities in hypersonic propulsion. So they set out to find the people with this experience, some of them already retired or moved away.

After finding the 46 people who had this specialized experience, Cadas set up a program that allowed younger, recently hired engineers to interview the older experts. The engineers met and went through a questioning process that enabled the younger employees to learn about supersonic technology and then to put that knowledge into a repository for future reference. "This was a double win for us," Cadas says. He adds that while in many cases older employees typically need some sort of incentive to participate, this time they were "falling over themselves" to teach the younger employees about the technology because they were proud of their accomplishments and were eager to see their knowledge retained for future use.

The experience at Rolls-Royce illustrates an urgent situation in which a cutting-edge technology risks becoming obsolete. But the death of certain skills isn't always a cause for mourning. For example, few would complain that they can't find a typewriter repairman in the phone book. However, organizations must carefully analyze whether a technology or skill might be needed in the future. "Companies need to ask themselves, What can we not afford to lose?" says Melissie Rumizen, senior knowledge strategist at consultancy SAIC and author of The Complete Idiot's Guide to Knowledge Management.

KM Can Aid in Succession Planning

Some companies turn to software to help predict future departures and determine crucial knowledge. Succession planning or talent management software can give organizations a good picture of who is working for them, how they are performing and how long they'll be around. With retirement on the horizon and new management positions to fill last year, automotive chain Pep Boys started using succession planning software from SuccessFactors to give it a clear picture of all employees at the company's 584 retail and service centres. "The risk of knowledge loss will always be there because there will always be unexpected departures," says Liviu Dedes, Pep Boys' director of training and organizational development. "But if you have a solid process to map out who is in your leadership pipeline, you'll be better prepared to fill job openings, retain top-performing employees and prepare for retirement."

Dedes says that the software lets him check how many employees are near retirement age and how many might be leaving soon for other reasons. Another way to get a visual picture of human interaction and pinpoint the go-to people in an organization is to do a social network analysis (SNA), which often involves interviewing employees and managers to see who is working with whom, and whom employees go to most often for help. SNA software can help organizations map out relationships and get a clear picture of who has the most knowledge and experience in a specific area (for more on SNA, see "Who Knows Whom, and Who Knows What?", CIO July 2005). The next step is to work on retaining those people, says Mohl at Children's Healthcare of Atlanta. "You need to focus on helping these people advance as quickly as possible," he says.