Why corporations don't always 'learn' their way to success

But a new study of the aircraft industry suggests that the belief in 'learning by doing' may be overstated. This study provides an alternative explanation for why we may observe a strong relationship between experience and performance, even without actual learning.

Successful companies engage in "astute self-selection" of what they do, said Jaideep Anand, co-author of the study and professor of strategy at The Ohio State University's Fisher College of Business. In other words, they find something they do well and simply repeat it.

"In some cases, companies don't learn so much from what they've done in the past as carefully choose to only repeat those activities with which they have been the most successful in the past and expect to be the most successful in the future," Anand said.

He argued that if a firm discovers that it has certain capabilities, it will repeat activities related to those capabilities, thereby simultaneously achieving more experience as well as superior performance.

"That is not the same as learning from experience," Anand said.

Anand conducted the study with Louis Mulotte of Tilburg University in The Netherlands and Charlotte Ren of the University of Pennsylvania. Their results appear online in the Strategic Management Journal and will be published in a future print edition.

Anand emphasized that he is not discounting the importance of learning from past experiences in all corporate activities. But he said there are situations in which learning from experience doesn't work.

Specifically, learning doesn't seem to help in corporate development activities such as new product introductions, diversification moves, international expansions, alliances, and acquisitions -- business activities that are infrequent, very different from each other, and where it is difficult to assess cause and effect.

Take, for instance, new product introductions. Companies don't normally put out many new products, these product launches are often very different from each other, and it is usually difficult to tell the reason why one particular product succeeded or failed.

"In any corporate activity where these three factors are in play, the role of learning from experience may be negligible," Anand said.

"There is a role of serendipity in this. If a company finds the right formula in the beginning, they can just keep repeating it."

The researchers tested their theory by analyzing essentially all jet aircraft projects launched by Asian, European, South American and North American firms between 1944 and 2000. The sample consisted of 437 new aircraft introductions undertaken by 159 firms. It included 189 fighter jets, 110 turboprops, 74 helicopters and 64 jets.

The data comes from the Aerospace Systems Group Library, which includes individual reports on each aircraft project commercialized since World War II.

Using a statistical analysis, the researchers concluded that firms tended to succeed by sticking with building aircraft in the same mode that offered them success in the first place.

"It is all very self-reinforcing. Companies repeatedly self-select the activities with which they have been the most successful in the past and expect to be the most successful in the future," Anand said.

There are a couple of different lessons here for managers, Anand said. First, don't overgeneralize from successes in activities like new product introductions that occur infrequently, are very different from each other and where the cause of success can't readily be determined.

"These are very complex activities and not easy to learn from," he said.

Also, if your firm is struggling with one of these complex activities, don't assume that you'll eventually figure it out.

"There's a danger in relying on learning in these particular situations. It is not always as successful as you believe," Anand said.