It doesn't take an affinity group to awaken most managers to the value of personal contacts. Who hasn't closed a deal -- from one end or the other -- that began with a call saying something like, "You don't know me, but I'm a friend of a friend"? That is networking at its most basic level; and it is on that most basic of levels that Allan A. Kennedy, for one, wants to marry human history to microprocessing.

"Networking is our primary distribution system," says Kennedy, a former McKinsey & Co. management consultant and co-author of the book Corporate Cultures, "yet it is depressingly inefficient. There's no good mechanism for retrieving [one's] history of personal relationships. Fortunately, that's just the sort of thing the computer's good at."

Kennedy, who now heads Selkirk Associates Inc., a Boston software company, has just sunk several hundred thousand dollars into the database architecture for a new product known as the Selkirk Networker. Designed as a kind of extended electronic Rolodex, the Networker program would receive and organize individual name files under an elaborate cross-referencing system. Its utility, however, is not as a stand-alone office system.

"The whole key is to get a critical mass of other Networkers," says Kennedy. "Once five or six plug into the same program, you begin to solve the 'small world' problem."

Which is?

"Academic studies have shown that with a statistical average of something like 2.7 movements, any transaction can make an enclosed loop around the country. In other words, by multiplying contacts by a relatively small factor, you can reach almost anyone in the United States on some sort of 'personal' basis. Dump half a dozen Rolodexes together and the potential is there to send networking's efficiency right through the roof."

Given its chain-letter premise, Kennedy admits he's "scared" to tackle the marketing of the Networker concept, an effort he figures could cost another $2 million. "Loading up [the data] can be extremely labor-intensive," he notes, "and it won't be worth it unless the customer sees the value of what he's getting in return. But it's a hell of an idea. With an initial threshold of 1,000 to 2,000 [customers], we'd have a critical mass. After that, you're talking about a mainstream product that a few hundred thousand people would have to have."