In the pre-PC era, IBM’s PR strategy was a conventional but effective “big company” approach that garnered IBM a great deal of public respect. The company invested in public charities, sponsorships of select TV and theater programs, advertising, and the usual editorial placements in a wide variety of publications to build and maintain its public image. Its approach was held up as a model of effective PR and marketing communications.
However, as was true of most IBM marketing programs, its PR program was highly centralized, not designed to communicate with its product marketing groups, and technically ignorant. This didn’t work well in the new era of press reviews and analysis that sprang up in the 1980s. Powerful columnists and influencers such as John Dvorak (an OS/2 fan) and Jerry Pournelle had little interest in IBM’s sponsorship of Hallmark’s annual showing of A Christmas Carol or its contributions to the United Way. But they were very interested in discussing the newest and hottest technology, playing with the latest technical toys, and having their egos stroked by people who were knowledgeable about the industry.
Over time, IBM developed an involuntary two-track approach to PC press relations. The first track consisted of IBM’s conventional PR program, which clanked along, oblivious to its increasing irrelevance in the new world.
The second track was an unruly back channel of former and current IBM employees who talked to the press on an ad-hoc basis, churning out gossip and fueling speculation. A mini-industry of “IBM watchers” sprang up, who were dedicated to deciphering the various statements and pronouncements of the different officers, divisions, and spokespeople.
Even worse was the fact that IBM had no formalized approach to managing its products’ review cycles, a problem that has plagued IBM since the release of the IBM PC and that continues to this day. Once an IBM software product is released, the product is on its own. Not surprisingly, very few IBM software products ever receive stellar reviews.
The first body blow to long-suffering OS/2 occurred when, in a major speech to business analysts, IBM CEO Lou Gerstner was quoted in August 1995 by the New York Times as saying that worrying about OSs was fighting the “last war.” Later in the speech, he added that it was too late for IBM to “go after the desktop.” Several newspapers immediately reported this speech as an admission by IBM that OS/2 was a failure. The New York Times article was headlined “IBM Chief Concedes OS/2 Has Lost Desktop War.”
The fallout was immediate and wide-ranging. OS/2 software vendors began to publicly question whether it made any sense to further invest in the OS. Many large corporate accounts that had committed to installing OS/2 on an enterprise level announced they were reconsidering their positions. Key advocates and columnists such as Will Zachman began to publicly question their support of OS/2.
The next, and even more devastating, blow came from a completely unexpected source. In his August 6, 1995, in the New York Times “Technology Column,” Peter Lewis ran a story called “OS/2 No Longer at Home at Home.” It was full of juicy quotes from an IBM spokesman. Among them: “OS/2 is a great operating system” but “Sony’s Betamax was a better system than VHS …” and “I’m going to put Windows 95 on the machines in my house.”
What made these quotes truly memorable was that the source was David Barnes, IBM’s Mr. OS/2 himself. Highly photogenic and comfortable in front of a crowd, Barnes had traveled thousands of miles over the previous 3 years conducting competitive demonstrations of OS/2 and Windows, had been a keynote speaker at trade shows, and had appeared on radio and TV extolling OS/2’s virtues. It was as if Bill Gates had been quoted as saying that Windows was really an inferior product to OS/2 and he wouldn’t be caught dead using the thing himself.
Reaction in the OS/2 community made the Gerstner faux pas seem insignificant. Online OS/2-friendly forums exploded. Tens of thousands of messages were posted electronically over the next several weeks, most asking for an explanation of Barnes’s remarks. Famous long-time OS/2 aficionado James Fallows, columnist for The Atlantic Monthly, former editor of U.S. News & World Report, and a noted writer, posted several public messages asking what on earth IBM was doing.
After the Barnes story broke, IBM did nothing for several weeks. Corrections weren’t published; Barnes didn’t write a letter of clarification to the editor of the New York Times; no IBM spokesperson appeared on any online services, Usenet forums, or SIGs to correct or explain Barnes’s statements.
Finally, after more OS/2 customers announced their defection from the product, IBM reacted. Barnes published a statement claiming that Lewis had taken his statements out of context. IBM assured everyone it was still committed to OS/2. Various IBM spokespeople made comforting noises. No one read any of these statements, and before Microsoft had even released Windows 95, its desktop OS, OS/2 was truly dead.
Cynics have pointed out that perhaps IBM was attempting to signal to the marketplace that it was discontinuing its support for OS/2. If this is true, it’s hard to imagine a more self-defeating strategy. At the very least, IBM could have waited until after Windows 95 had shipped to judge market response. But after the Gerstner/Barnes remarks, Microsoft could have waited another year to release Windows 95. It wouldn’t have mattered.
As already noted, OS/2’s failure not only had a profound impact on IBM, but it also altered the fate of many companies in the industry. Perhaps the saddest case was that of SPC, the firm that through no fault of its own had years ago inadvertently yoked me to TopView for that one miserable day. SPC had divested itself of its PFS line by the late 1980s and, via its purchase of Harvard Graphics, was for a brief period the leader in the PC presentation graphics market. SPC made a “bet-the-company” wager on IBM and OS/2 and developed InfoAlliance, a high-end OS/2 database product. Such was SPC’s confidence in OS/2’s future that it literally ordered its sales force to cease selling its market-leading Harvard Graphics package and concentrate on InfoAlliance. Whoops. When it became clear this bet wasn’t going to pay off, SPC turned around and spent 2 years rewriting the package for Windows, but by the time the project was done, the company was out of cash and market share.
Even those companies who avoided being sucked into OS/2 development efforts ended up paying the price. Many mistook the market’s failure to adopt OS/2 as a repudiation of GUIs. They had several clues this wasn’t true—the enthusiastic reaction to Microsoft Excel and Word for Windows being but two examples. Apple’s Macintosh success, despite the fact the company was held in distaste by much of corporate IT, was another. But such was IBM’s hold on the market’s perception that many believed that DOS would remain supreme on PCs for several more years.
Had companies such as Borland, Lotus, and WordPerfect committed to Windows development efforts in 1988 or 1989, they would have been in a position to compete with Microsoft on a fairly level playing field when Windows 3.0 took the market by storm in 1990 and 1991. The opportunity was certainly there; during this time period Microsoft was desperate to garner ISV support and went to great lengths to court potential developers.
Much ink would be spilled in the mid-1990s over Microsoft’s creation of “secret” API calls that supposedly gave it an unfair advantage over its rivals. Most of this was nonsense. What hurt these companies was not code but time—the time they had to take to play catch-up with Microsoft, which was ready to release what the market wanted: robust, GUI-based Windows products.
Even more ink has been spilled in bemoaning Microsoft’s supposed perfidy in taking advantage of poor old trusting IBM in their joint venture to bring OS/2 to market. Such sympathy is wasted. IBM bears almost complete and direct responsibility for the failure of OS/2. From advertising and pricing through to positioning and naming, it’s difficult to find a marketing mistake IBM didn’t make. The truth is that by 1990, the PC market was ready to accept almost any GUI-based system that worked, and Microsoft simply provided what everyone wanted. Bill Gates is undoubtedly a very smart guy, but someone with half his brains could have whipped IBM.
What ailed IBM then and what ails it today is that the company is simply too big. Although no one has ever been able to identify exactly when a company becomes so huge it can no longer effectively compete, by the early 1990s IBM had clearly reached that point. With 400,000+ employees, and products that competed in every segment of the market in almost every country of any note, IBM of necessity had to manage and assign priorities to a welter of competing interests and initiatives. It was a task of dizzying complexity and perhaps a business genius could have managed it. Large organizations, however, tend not to promote geniuses to top managerial positions. Geniuses tend to be monomaniacal in their focus, less than solicitous of other people’s feelings, and often make those around them uneasy (a description that reminds many of Bill Gates). Smart politicians are the types who usually climb to the pinnacle of corporate success in large companies, but a company of IBM’s size needs more than an affable organization man to kick it in a desired direction. Though during his tenure at IBM, Lou Gerstner was lauded by the press for the company’s modest turnaround, his financial accomplishments came more from cost cutting and retrenchment than renewed business growth. And shortly after Gerstner’s departure, his successor announced that, yes, things were still rather slow at IBM and more layoffs would be coming.
The answer to IBM’s problem, ironically, was discovered by the U.S. government in the 1970s when it attempted to break IBM up. IBM fought the government tooth and nail and eventually prevailed, allowing the company to remain an increasingly unresponsive muscle-bound giant unable to get out of its own way. By the early 1990s, John Akers decided the government had been right after all and developed a plan to split the company into several autonomous divisions. Akers was shown the door before he could put his plan in motion and IBM remained intact, but these were Pyrrhic victories. Voluntarily, of its own accord, IBM should break itself up into different companies and allow each to pursue its own destiny. Some will fail, but the units that succeed will restore growth and vitality to individual businesses that now, collectively, have very little.