IBM Should Use Japan’s Formula
IBM Chairman John F. Akers has pulled a Gorbachev. The world’s largest computer company is taking a $3-billion writedown, trims yet another 20,000 workers from the rolls and launches a “fundamental redefinition of the way IBM does business.”
Instead of managing itself as a monolith, Akers announced Tuesday that IBM will break itself into a confederation of quasi-autonomous units--a high-tech network of multibillion-dollar IBM-lettes, each capable of more nimble and profitable responses to the technologies and markets that drive them.
But perestroika doesn’t necessarily mean profit. The redefinition IBM really needs to succeed is Made In Japan. Indeed, IBM’s best hope for organizational renewal can be found in the design of Japan’s enormously successful and innovative zaibatsu and keiretsu : the Mitsuis, the Sumitomos, the Mitsubishis.
These industrial combines are huge, global, rich in culture and tradition yet immensely adaptable. They offer models that, with diligence, thought and creativity, should be emulated--not imitated. IBM needs to become America’s high-tech zaibatsu of the 1990s.
Instead of setting up a holding company structure with quasi-independent business units, IBM should take a tip from Japan and assure that there are cross-holdings between the various businesses. Management influence should come horizontally and diagonally--not primarily from on high.
To wit, IBM Mainframe should have a piece of IBM Personal Computers. IBM Personal Computers should have a stake in IBM Networks. All of the IBM companies should have a stake in IBM Silicon/Semiconductors. And, of course, there should be an IBM Standards Steering Committee that debates and arbitrates software and telecommunications standards for the entire company.
Indeed, the zaibatsu/keiretsu structure creates all sorts of interesting opportunities. Maybe IBM Personal Computers would like to invite Microsoft as an investor. Perhaps IBM Silicon/Semiconductors feels that a joint venture with Motorola makes sense. Suppose Mitsubishi would like a piece of IBM Mainframes? By creating keiretsu -like networks of cross-holdings and collaborations, it becomes safer for the individual IBM units to create relationships with outside companies.
As the Japanese experience demonstrates, that’s a key to successful technology transfer from the outside.
“Basically, these groups came into being in the Meiji era as brilliant devices for the rapid importation of advanced technology into a country with scarce capital,” says Chalmers Johnson, a professor of Pacific international relations at UC San Diego and a leading scholar of Japanese industry. “They are unbelievably adaptable and, serendipitously, a perfect fit for high-technology industries where the life cycles are so short and the R&D; costs are so high.”
Indeed, Johnson asserts, this “developmental conglomerate” structure not only offers the benefits of diversification but also, if the relationships are smoothly handled, provides opportunities for creative interactions.
“The zaibatsu promotes the combining of utterly diverse technologies; what the Japanese call ‘fusion research,’ ” he says. “The very nature of the organization produces collaborations between utterly diverse parts of the organization.”
That sounds a lot like what IBM needs. Simply granting independence and autonomy to a dozen IBM-lettes is not the way to build a cohesive organization. Does IBM really want to run the risk of turning itself into an information technologies conglomerate where managing for shareholder value takes precedence over providing computational options for customers?
John Akers ain’t a conglomerateur like ITT’s Harold Geneen. IBM’s top management doesn’t know how to run the corporation as a portfolio. As the folks in the Soviet Uni--excuse me, Russia and Eastern Europe--have shown the world, when you fundamentally restructure an economy (and IBM is a global economy!), you unleash forces you can’t quite control and can’t quite understand.
No question, IBM has to be dramatically restructured. But the real challenge isn’t figuring out the best way to draw the dotted lines and divvy up the company. It’s figuring out what are the themes, standards and values that bind everyone together. That’s what Japan’s zaibatsu/keiretsu are so good at managing.
The zaibatsu/keiretsu experience unambiguously proves that industrial organizations can effectively balance size and flexibility as well as innovation and profit. The key is not to grant autonomy but to build intelligent interdependence between the appropriate individual units. In other words, create market incentives for cooperation, coordination and collaboration.
If IBM’s top management has redefined the world’s most important computer company along these dimensions, there is every reason to believe that its investors, employees and customers have a lot to look forward to. If it has not, we’ll be reading about “IBM’s Yeltsin” in the business press before the end of next year.
