Bill Sanko had good news when he met with his parent company, GTE Corp., in May 1984. The autonomous division he had spent four years creating was just turning profitable. He had used less money and won more customers than GTE had anticipated. And repeat orders were piling up.

But GTE had only one thing to tell Sanko: It didn't want the division anymore. Period. "There was a sense of disappointment for us," he remembers." All of us had put a lot into it."

Many big companies like GTE are setting up separate business units to pursue promising ideas. They free these units from existing divisions so that the new venture doesn't have to compete with an ongoing business for resources and attention. But there are still pitfalls that can make even successful new divisions unwelcome.

Corporate changes -- a company reorganization, a management shift, or growing impatience -- can reshape a thriving venture into the proverbial square peg. GTE decided to consolidate and focus on such bigger markets as communications-switching equipment. Sanko's niche in telephone-transmission products didn't fit.

Management changes can also hurt. After shifts at Rockwell International Corp. in 1982, its textile machinery division sensed -- but was never told -- that it didn't belong anymore." They still required us to come to headquarters every month, but we were denied R&D funds," says Ronald Brown, marketing vice-president of the venture, now a private concern called Draper Corp. "None of us knew what was in Rockwell's thinking," he says.

Parent companies also tend to underestimate the time it takes to develop a new business. "It's a management attention-span problem," says Michael Brose, president of Technology Consulting Group Inc., a Boston management consulting firm. "They pull the plug because they don't have the time to manage it or nurture it." One consultant estimates that the average life span of a new venture within a big company is four years, about half the time it needs to prove itself.

Setting a unit free, though, can pay off for both sides. A division with solid assets and strong management can put together a leveraged buyout. Given the time, it can negotiate with potential investors and acquirers. GTE gave Sanko and his key managers 90 days to find new investors, venture capitalists now own half of the private company, XEL Communications Inc. "Being up front about it is important," says Sam Felton, director of the Midwest region for SRI International. "If you ignore the division, or if you starve it because it doesn't fit, you are headed for trouble and, probably, a big write-down."