Chapter 20 - CEO Tumbles

Boards of directors have several duties, notably providing transparent assurance that the processes for financial reporting, executive compensation, and corporate governance meet the requirements of investors and regulators. But their most important responsibilities include firing the failing CEO and recruiting a replacement who can successfully lead the company for the next ten years or so. On these issues, the record of the HP Board is questionable at best.

Despite the company’s placid appearance, top management was caught up in continuing turmoil, caused, in part, by constant interference from the founders (and their surrogate) for years after their formal retirement. Paul Ely, who hoisted HP into the computer business, was fired for being “abrasive,” with the active connivance of John Young, the CEO. Young was forcibly but tactfully retired by the founders for bureaucratic centrism. His successor Lew Platt was ousted by the founders’ surrogate. The flaw? Not abrasive or centrist enough. And the successor’s successor, Carly Fiorina, was fired as being too removed from the details.

Strategic coherence and consistency suffered. Until quite recently, almost all of HP’s profits came from printers and toner, the one segment of the business that was shielded from the continual perturbation. This isn’t an argument for blind CEO retention, but HP’s twenty-five painful years do suggest Board caution. Naomi and I were aghast at Paul’s sudden downfall. There were unverified stories of board-level politicking and intrigue. Paul had been “too abrasive” for company cohesion, relayed Ed Matthews, an RB founder with strong ties to HP. Yet here was a man who had headed an organization with arguably the strongest management team in the computer sector at the time. Under his leadership, the company’s nearly disastrous HP 3000 program was salvaged, its computer revenues soared to $3 billion, and substantial progress was made towards a RISC-based computer generation and Unix.

Young launched the new computer on a trifecta gamble— simultaneously changing hardware architecture, operating system, and management team. This seemed even riskier than IBM’s attempt to change hardware and operating software on the “bet-the-company” System/360 or the failed Future System.

But those lessons were apparently lost on HP’s leadership. John Young was pushing a two-way gamble into a three-way improbability, especially once he widened R&D’s purview to embrace measurement and instrumentation as well as computing. The complexity of managing IT development by committee would yield terrible results.

In 1985, we interviewed Young, COO Dean Morton, and John Doyle, Ely’s replacement as executive VP for information systems and networks. All three were smart, first-rate executives. Doyle accompanied us on our tour of HP facilities in Idaho and Colorado. His disparagement of Paul Ely, though infrequent and indirect, was nevertheless unmistakable. “We can once again disagree without being disagreeable,” he told us. “Or agree without the terminal niceness that merely signals suppressed rebellion.” The disagreement apparently stemmed from the fact that the company’s best computer scientists gravitated toward Birnbaum’s research activity, while the traditional instrumentation and medical businesses were steadily losing ground to niche competitors with better computing capability.

John Young was a healthy, Nordic-looking man, whose occasional body flexing made him seem curiously self-conscious. From his perspective, the most significant event of the past eighteen months was the reorganization repositioning the company “from the product side to the market side.” No longer was there a split between computing and HP’s traditional products; now everything was reoriented around six market segments—measurement, engineering, medical, analytical, information processing (IT), and manufacturing. All HP business units would share equally in the computing advances engineered by Birnbaum’s organization.