When innovation supersedes short-term results

It can be the thorniest passage for any business that aims to be around for the long haul: readying the next generation of products or services while effectively managing the ones that currently generate sales and profits.

There is no one-size-fits-all approach. But some captains of commerce have begun to rethink how they evaluate the innovative programs being incubated under their companies' roofs. And one suggests that, for emerging ventures, it may be more important to measure learning than to focus on short-term financial results.

''The challenge is to balance a culture of accountability in established businesses with a culture of learning in experimental businesses," Ray Stata, founder and chairman of Analog Devices Inc. in Norwood, wrote in the spring issue of MIT Sloan Management Review.

Stata believes the stakes could not be higher at a time when the life expectancy of Fortune 500 companies is 40 to 50 years, and falling. In an age of globalization and rapidly changing technology, every chief executive will need to transition in and out of businesses and markets.

It is seldom an easy, or graceful, process -- and business graveyards are cluttered with the companies that failed to make the necessary transitions. But Stata has come to believe that constraints on innovation have less to do with deficits in creativity or technology than with the capability of management to foster new businesses.

''You have to take resources away from your core business and put it in more speculative businesses at a time when you're really flying," Stata said in an interview. ''It can be hard to justify."

And when financial results fall short of targets, companies are programmed to conclude that managers underperformed. For emerging businesses, an alternative explanation might be that the forecasts were overly optimistic and need to be revised on the basis of experience. It can be sorely tempting to abandon pilot programs slow to bear fruit, but the best long-term move may be a decision to stay the course.

Charles A. O'Reilly, professor of management at Stanford Graduate School of Business, is collaborating with Harvard Business School professor Michael L. Tushman on a book exploring the practices of ''ambidextrous organizations." Such organizations set up separate benchmarks for mature businesses, high-growth businesses, and future businesses.

In the case of IBM Corp., for instance, that would mean measuring its core computer hardware business differently than its service and consulting business or its new life sciences initiatives.

''Being ambidextrous means accepting the fact that you may have some businesses where the metrics and competencies and alignments are going to be different from other businesses, and not superimposing the same way of managing," O'Reilly said.

Analog, a semiconductor company with more than $2 billion in annual sales, has reinvented itself several times during its nearly 40-year history. The company branched into digital signal processing in the 1980s, and microelectromechanical systems (called MEMS) in the 1990s, and it's now backing a team experimenting with phototonics.

''In every case," Stata acknowledged, ''it's taken well into a decade before you knew you'd made the right choice."

The idea to develop MEMS technology for automotive air bag sensors came from a group of engineers at Analog's semiconductor division in Wilmington. The team drove the research to the point where it began to siphon off resources from the core business. ''The conflicts escalated, and finally we had to pull the operation out of Wilmington and establish a standalone division in Cambridge," Stata recalled.

Some top managers and directors sought to pull the plug, but Stata determined ''we had to go through the learning curve, and it was taking more time than expected."

The business today is profitable and has become the leading supplier of sensors to the auto industry.

That experience yielded some lessons, which Stata shared in his article in the Sloan Management Review.

''Companies need a system that helps them continuously refine judgments regarding the likelihood and timing of success," he wrote, ''and also accelerates learning so that failing strategies can be quickly modified."

Robert Weisman can be reached at weisman@globe.com.