AT&T;’s Kavner Trying to Fill Some Big Shoes

Last month AT&T;’s computer group head, Robert M. Kavner, spent a few weeks vacationing on the Italian Riviera with his corporate predecessor, the suave and ultra-stylish Vittorio Cassoni. Although golf was the major recreation, Cassoni insisted on taking Kavner on a “major money” shopping expedition for his fall wardrobe.

Ever since, the reserved Kavner, an accountant by training, has been wearing the same style of double-breasted Italian business suits that the charismatic Cassoni favors. But an important remains: Can Kavner can fill the Gucci loafers that Cassoni left behind last April when he quit AT&T;?

The answer--which could well determine whether AT&T; will ever become the major force in the computer industry that once was so widely expected--has yet to emerge. And it may be many more months before Kavner will be able to focus all of his energies on executing a winning computer strategy for AT&T;, which has already lost an estimated $2 billion on its computer business since jumping into the field in 1984.

AT&T; had been betting on Cassoni, a witty engineer with considerable marketing savvy, to lead the company’s computer operation out of the doldrums. When he left abruptly to return to Italy and a job at Olivetti, the conventional wisdom was that AT&T; would lose the momentum Cassoni had built up. With the selection of Kavner, then AT&T;’s chief financial officer and a virtual unknown in the computer industry, the pessimism grew. “What can a bean counter know about computers?” asked one analyst at the time.

Clearly, Kavner, a spry, 44-year-old marathon racer, has not enjoyed an easy introduction into the computer industry.

Since becoming president of AT&T;’s Data Systems Group, he has been forced to devote most of his attention to a computer industry feud that started just weeks after his appointment. Known within the industry as “the Unix wars,” the fight has pitted AT&T; against such formidable foes as IBM and Digital Equipment Corp.

At issue is the future development and direction of the Unix computer operating system that AT&T; invented nearly two decades ago, a system that broke new ground in the industry with its ability to link disparate types and brands of computer equipment. The two sides are battling over which version of Unix will emerge as the industry standard. AT&T; had hoped to get a leg up on its competitors, and win new customers, by establishing a version of Unix it is developing with Sun Microsystems as the dominant one in the computer industry.

According to several industry sources, a settlement may emerge within weeks from the tense negotiations, which now are said to be at an extremely delicate and crucial stage. If so, insiders say it will a be a tribute to Kavner’s deft negotiating skills.

“Kavner is more of a negotiator than a posturer,” says one financial analyst. Says another industry insider: “Bob doesn’t let his own ego get in the way. He’s results-oriented and he looks at the bigger picture.”

Strategy Uncertain

Although AT&T;’s poor computer showing to date is now Kavner’s problem, it was hardly of his making. It has stemmed, analysts say, from AT&T;’s failure to develop the kind of vision for its line of general business computers that has long guided its vast telecommunications operations.

“If AT&T; has a computer strategy, I’d sure like to know what it is,” laughs Paul Cubbage, an analyst with the Dataquest market research company in Silicon Valley. “They said in 1984 that they wanted to be a significant computer player. Now I figure that means that they have a lot of money that they’re going to spend on the computer business. . . . All they’ve done is develop some products and shove them on the market.”

Indeed, AT&T;’s dismal computer sales have fallen far short of the company’s original goal of seriously challenging those of IBM and the other industry leaders.

Last year, industry insiders say, the company sold a relatively meager 120,000 personal computers and fewer than 8,000 minicomputers. AT&T; has said, without elaborating, that it expects to meet its goal for this year, which some sources have described as a 30% sales increase over 1987. However, industry analysts point to the company’s recent round of price cuts as a possible sign of sluggish summer sales.

Kavner says the heart of AT&T;’s computer strategy is to take advantage of the company’s telecommunications expertise, which will help its customers tie their computer systems together in networks that share information speedily and efficiently.

“We have a particular strength in telecommunications,” Kavner says. “Our business is not just selling boxes. It is to provide the ability for offices to have access to networks.”

But with most of the rest of the computer industry talking about networks, too--1988 has been called in some quarters the “Year of the Network”--it’s still difficult for many analysts to determine just how AT&T; intends to distinguish itself from the pack.

One clue may be in the alliance established by AT&T; this year with hot, young Sun Microsystems, a Silicon Valley maker of sophisticated engineering and technical workstations. The deal, which calls for AT&T; to buy up to 20% of Sun, was widely viewed as an indication that the lumbering and bureaucratic AT&T; wanted to change its image by teaming with one of the slickest and fastest-moving operators in the industry.

Analysts say the alliance with Sun should help AT&T;’s stagnant product line. “You know what they always say about AT&T;,” says Maria Sbrilli Lewis, an analyst at Shearson Lehman Hutton in New York. “Long on R (research) and short on D (development).”

Even before AT&T; agreed to invest in Sun, the two companies had been working to develop a standard version of the Unix operating system.

Originally written at the Bell Laboratories in 1969, Unix is one of the very few computer operating systems that works on a variety of machines. Thus, it allows customers to mix and match their equipment without worrying about whether the machines can share information. Under Unix, they can. And customers love that flexibility.

Many Versions

However, Unix has long suffered from being a creation of AT&T.; Until it divested its telephone operating companies in 1984, AT&T; was prevented from entering the computer business. Unable to profit from Unix, AT&T; virtually gave away the operating system, which controls the flow of data in computers, to universities and research groups where it blossomed in a multitude of versions.

Many of those versions, such as the one developed by student hackers at the University of California at Berkeley and now used by Sun Microsystems, became the basis for highly profitable computer companies. However, with so many versions, applications software writers had a hard time selecting the right one on which to base their programs. As a result, Unix has yet to develop the rich library of applications software--the programs that actually perform tasks for computer users--that other operating systems have.

That’s where the notion of a standard Unix comes in. The version Sun and AT&T; wanted to develop and use in their machines was supposed to have been based on their own Unix specifications. But dozens of competitors, including IBM and Digital Equipment, protested that the AT&T-Sun; effort could push other manufacturers of Unix machines out of the market.

The challenge prompted AT&T; to rethink its Unix strategy. Over the last several months, the company has slowly backed away from its original position that it would work only with Sun on the new version. In fact, there is widespread speculation that AT&T; may join with the IBM-Digital group to write another standard version of Unix.

“It’s been somewhat of a humbling experience for AT&T;,” says Lewis, the Shearson Lehman analyst.

AT&T; officials won’t comment on the negotiations, but insiders report that several points need to be settled. One potential outcome is the spinoff of AT&T;’s Unix operations into a separate business unit. Analysts say the company already collects about $40 million in Unix royalties, an amount that would make that business the nation’s 10th-largest software company. Kavner has already endorsed some sort of Unix spinoff.

An end to the skirmishing would allow Kavner, who has been intimately involved with the talks, to devote more attention to his division’s sales. It also would give him a chance to demonstrate his leadership skills. Although Kavner is widely regarded as extremely bright and hard-working, the appointment of an accountant to head a computer company raised more than a few eyebrows.

Differences Downplayed

Insiders say, however, that Kavner has surprised some by holding his own with engineers and other “propeller heads,” as the technical whizzes are often called.

For his part, Kavner says he never doubted he could handle the job. When others were referring to him as “AT&T;’s bean counter,” Kavner pointed to his long-standing personal and professional relationship with Cassoni, a friendship that he says provided the basis for his understanding of computers and where AT&T; wants to go in that industry.

In fact, Kavner says he’s really not very different from his predecessor. “We were bachelors together in New York,” he says. “We both married at the same time and we both have a young child now. Our wives are close friends.”

Although Cassoni now lives in Italy, Kavner says they talk on the phone a couple of times a day, although the conversations are almost exclusively about business. “Cassoni and Kavner are best friends,” he says.

And the two share the same vision for AT&T;, Kavner says. “If you were talking to Vittorio now, you would would be hearing everything I’ve said but with an Italian accent.”