(FORTUNE Magazine) – THE VOICE was Tennessee drawl, but the sentences were clipped and pointed: ''There is going to be change. This is a tough board. This is not American Express.'' The speaker was an IBM director, Thomas Frist, chairman of Hospital Corp. of America in Nashville. He was talking on the evening of January 26, only hours after the end of a long IBM board meeting, after Chairman John Akers had announced he would step down and after the directors had officially launched a search for a new chief executive. One day before, the American Express board, supposedly on the way to tossing out Chairman James Robinson III, had made itself a laughingstock by letting him stay -- changing his job, yes, but leaving the business world with the impression that the deck chairs had been rearranged, nothing more. Robinson subsequently departed but, as Frist avowed, a Blue-tinted replay of boardroom weakness won't happen at IBM. After eight tumultuous years at the top, the last one scarred by a $5 billion loss, Akers, 58, is headed out. In FORTUNE's opinion -- though the board does not concede its thinking has gone this far -- the new CEO will be an outsider, brought in to shake the place to its foundations. Leading the urgent search, set to take no more than three months, is director James E. Burke, 67, retired chairman of Johnson & Johnson. At his side is his close friend and fellow director, Thomas S. Murphy, 67, chairman of Capital Cities/ABC. More than any others, these two forceful, class-act executives are apt to determine which Mr. X will next run this troubled company. John Akers took the graceful, even heroic, way out by saying he would move on, thereby opening the way for the radical change in leadership the company needs. Akers, though, really had no choice. At this point, the board would not have let him stay. He had lost his constituencies, among them IBM's shareholders, who have watched $77 billion in market value disappear since the top in 1987. Institutional shareholders were clamorously demanding change. Distraught individual investors were bombarding IBM headquarters in Armonk, New York, with calls and letters. One mid-December missive to Akers, sent by a New York shareholder who said he had owned the stock for 40 years, testily cited Oliver Cromwell's protest to the Long Parliament: ''Begone, you have sate long enough. Do you think to sit here till Doomsday come?'' With IBM's dividend just cut by more than half and the stock recently below $50 a share, many an IBM shareholder thinks Doomsday is already here. The company's considerable strengths -- notably what Jim Burke calls its ''superb product and technology base'' -- are undercut by weakened demand for mainframes, its core competence. They are enveloped by a thick batting of Big Blue bureaucracy. The first burning question about IBM today is who will be selected -- who, indeed, can be persuaded -- to take on the monumental job of running the company. The second question: Can the new boss unshroud the good things about IBM and bury the bad? Despite years of trying, Akers barely managed to rip a fingernail through the shroud. For eons, though, he kept the confidence of a board that admired the resolve with which he had tackled IBM's almost unremitting problems. Quizzed by the press, IBM's directors invariably stuck up for Akers, even after board revolutions at General Motors and other large companies in 1992 signaled that relationships between directors and management were irreversibly hardening. The board's support for Akers finally began unraveling in December. By then, bleak business in Europe and Japan had turned IBM's third quarter into a surprise disaster and was wrecking the fourth, normally the best three months of the year. BEHIND THE SCENES, the five nonmanagement directors on the executive committee -- that is, everybody but Akers and President Jack D. Kuehler -- began to talk the heretofore untalkable: changes in management. The committee's leaders were Burke, Murphy, and John R. Opel, Akers's predecessor as CEO. A step down in influence were the committee's other members, J. Richard Munro, retired chairman of Time Warner, and Stephen D. Bechtel Jr., chairman emeritus of Bechtel Group. The rest of the board was largely out of the loop. Signs of roiling change appeared immediately after a special board meeting in mid-December, called to confront the worsening financial situation. Management appeared before security analysts and the press to acknowledge that the $4.84 annual dividend, which cost the company more than $2.7 billion annually, was no longer inviolate. Days later, the stunning news arrived that a retired top IBM executive, Paul J. Rizzo, 65, would return as a consultant. In the early 1980s, Rizzo, a seasoned financial man, had been Akers's rival for the job of CEO and the candidate John Opel most favored. But Akers had the strong backing of another IBM power, Frank T. Cary, CEO from 1973 to 1981. Just turning 50, Akers was also the right age to have a long spell as boss before reaching IBM's traditional retirement age of 60. Rizzo was called back to provide heavyweight help for a company beleaguered by missed earnings forecasts, never-ending write-offs from downsizing, and sudden financial craters. ''Fact is,'' Jim Burke said in late January, ''we weren't getting good information. Some of that is explicable because things were happening so quickly in the business ((the slump in Europe, for example)) and nobody had experienced it before. They didn't know how to quantify it the way it needed to be.'' But another part of the problem, many Wall Streeters think, was weak financial management by chief financial officer Frank A. Metz Jr., 59 -- a marketing man who'd made the jump to finance, the nicest guy going by everybody's agreement, but an executive not really a match for IBM's vastness and complexity. Even the basic numbers are overpowering. This is a company that commands $65 billion in revenues, carries $30 billion in debt, and plows some $10 billion annually into capital expenditures and research and development. Up against that, Metz appears often to have floundered. Eventually, he also exhausted the patience of the directors. The board meeting in the last week of January turned into a clean sweep, opening up all at once the slots of CEO, CFO, and president. Rizzo, newly elevated to vice chairman, takes over for Metz, who retires early. Kuehler, 60, IBM's technology guru and president, had stayed past customary retirement age to help IBM cope with its crisis; he will continue to do that, but as another vice chairman. Burke says Rizzo is not a candidate for CEO and will probably leave IBM within, say, two years. In the meantime, he would step aside were the right CFO to be found. The three top spots could be filled one by one -- the second two possibly by a new CEO installing his own key executives. Heading a seven-person search team made up of his friend Murphy and five other outside directors, Burke can be expected to solicit all opinions as to who the next leader should be. In FORTUNE's view, remarkable, creative solutions are not likely to be ruled out. IBM could well trade on the widespread belief that it is a national asset and ask for special dispensation from Washington to make a merger that would otherwise never be allowed. How do you like the idea of IBM Apple Corp., headed by John Sculley? Or Microsoft IBM, run by that 37-year-old upstart -- you know who? Opinions about the type of boss IBM needs and what he -- or she, but why does that word never arise? -- must do inevitably reflect people's views of what IBM now is not. Ram Charan, a one-man-band management consultant who has advised such companies as General Electric, Citicorp, and Conrail, paints a not atypical picture: ''First, he's got to develop credibility with customers, employees, and markets. He has to have a demonstrated track record and be able to sort out what needs to be done with this complex company that has a lot of things intertwined like spaghetti.'' Total junking of IBM's famous full employment policy, already eroding, will probably be necessary. The company is down from 406,000 to 300,000 employees, but many outsiders argue it will have to go lower, fast. Computer executives mention cuts on the order of 50,000 or 75,000, typically adding that there's no science to their estimates. Downsizing of that magnitude would cost IBM dearly. In December it confirmed an analyst's calculation that it was charging about $125,000 to earnings for each employee headed out the door. The money goes for rich special retirement plans and ordinary severance pay. At that rate, to cut 75,000 employees would require a charge against earnings of more than $9 billion. For perspective, the company's net worth is now around $27 billion and heading south on an out-of-control train. Benefit plans, however, are not immutable. It's not hard to imagine a new CEO deciding that the company cannot bear all the costs that it has been doggedly accepting as part of its heritage. So what should IBM do now? A lot of what Akers was trying to do -- too slowly. An agenda for Mr. X:

SET THE BUSINESSES FREE -- The sum of the parts at IBM is greater than the whole. Management must immediately identify its most self-sustaining businesses and give them full autonomy, selling off chunks in many cases. Vertical integration worked well in the mainframe era, but that's history. Today, corporate alliances constantly shift and evolve. Managers must continually decide where, how, and with whom to design, build, and sell their products and services. Those decisions must be made for the benefit of particular businesses, unconstrained by worries about a larger whole. Akers claims that most of IBM's businesses are being freed to compete on their merits, even if salesmen from one stumble over IBMers from another in customers' waiting rooms. But this is largely a fallacy. Three units are supposedly moving quickly toward spinoff -- the IBM PC Co., Pennant Systems (printers), and Adstar (disk drives and other storage devices). Yet none fully controls its product distribution or advertising budget. A single IBM sales organization handles the whole range of products, from $99 PC software to $20 million mainframes. ''One sales force is absurd,'' says Charles Morris, a consultant and coauthor of Computer Wars, which analyzes IBM's ills. ''These are totally different businesses, and the proper distribution system for each will be quite different.'' The incoming CEO ought to ponder the example of Lexmark International, which took over IBM's consumer printer, keyboard, and typewriter businesses. Since a leveraged buyout less than two years ago -- IBM retained 10% -- it has consistently exceeded projections. Profits rose 50% in 1992. Lexmark freely acknowledges it couldn't have thrived without having its own dedicated sales force for the first time.

GO MICRO -- IBM says it is already doing this, but it's at least five years late and insufficiently committed. Today's customers want Lego-like systems, networks of microprocessor-based PCs and workstations that can be expanded and modified cheaply. They also need to perform complex tasks that used to require mainframes. Who better to help them than IBM? But IBM wouldn't, so they turned to others. One of IBM's most successful products, known as CICS and pronounced ''kicks,'' is complicated software that manages transactions in big systems like those used by airlines and insurance companies. Not until last year did IBM make CICS available on workstations -- and then only after competitors brought their own versions to market. It's easy to see why IBM resisted shifting from a mainframe mentality. Even as they decline, mainframes and their svelter cousins, minicomputers, are by far the most profitable parts of its business. Two reasons: Gross profit margins are much higher -- 60% and up, vs. 30% to 50% for smaller systems -- and big iron sales typically bring with them an annuity in the form of multiyear software and maintenance contracts. Says Todd Hixon, senior high- tech consultant at the Boston Consulting Group: ''That's a kind of financial cocaine. It's very hard to get off.'' But if IBM doesn't cannibalize its own businesses by offering cheaper technologies as soon as they're available, others will. IBM has to walk a tightrope, trying to maximize profits in the face of advancing technology. A case in point: the company's forthcoming parallel processing machines, which could take business away from IBM's costlier mainframes.

MARKET BY INDUSTRY -- Once IBM products have their own sales forces, the rest of the marketing staff must be reconfigured. What's urgently needed by business customers are teams of systems integrators organized around specific industries -- experts who know more about, say, banking technology than the bankers themselves. That sort of structure has allowed EDS, IBM's most successful competitor in computer services, to double annual sales to about $8 billion in five years. IBM's systems integrators must be allowed to supply whatever serves the customer best, whether it comes from IBM or elsewhere. And they should provide lots of custom, industry-specific software. Says Bob Djurdjevic of Phoenix, a consultant whose Annex Research follows IBM closely: ''If services are the way to go, then you must organize yourself around the customer.'' Such a change would not be wholly revolutionary for IBM. Its fast-growing U.S. subsidiary known as Integrated Systems Solutions Corp., which assembles and operates computer systems for companies, is already largely organized by industry. So are a number of European marketing groups.

KEEP SLASHING COSTS -- Along with the continuing loss of jobs must come more asset write-offs. Djurdjevic notes that IBM sank $42 billion into plant and equipment in the 1980s, most of it related to mainframes that are increasingly obsolete. ''The issue is downsizing in costs,'' he says, ''and people costs are just one element. The company has done a much better job of reducing employment than writing off unnecessary physical assets.'' For example, Djurdjevic says, Adstar, IBM's storage products operation, could probably close some of the ten factories it maintains around the world. If a consistently profitable IBM is to emerge, it will be not only broken up into more competitive pieces but also smaller overall. In fact, IBM may be genuinely unmanageable on its present scale. Too many disparate and competing businesses contend for capital and top management's time.

THROW OPEN THE WINDOWS -- Probably nothing in IBM's corporate culture does more harm than its arrogant and secretive style. IBMers yearn for a sense of common purpose. If a vigorous new enterprise is to emerge, its leader must articulate a fresh vision, rallying the troops around new ways of serving the world with technology. The personality of the next CEO will be crucial. Arrogance no longer sells, internally or externally.

BLAST AWAY AT THE BUREAUCRACY -- Finis Conner, chairman of Conner Peripherals, a competitor of IBM's in disk drives, zeroes in on what is probably Big Blue's biggest problem -- its impenetrable bureaucracy. ''Companies build up plaque,'' Conner says. ''They need plaque busters.'' At IBM, a few won't do. This company needs a whole dental school. An insight into what it's like in the real world, as opposed to IBM's, came recently to Joseph Guglielmi, 51, who left the company last year after 30 years. He now heads Taligent, an IBM-Apple joint venture that is working on ways to make it easier to create complex PC software. At IBM, Guglielmi says, it was standard procedure for executives' assistants to sweat the details, freeing up their bosses to ponder ''macro'' issues. That shields the bosses from the outside world: ''It's harder to keep in touch with the market and customers and what's going on.'' Traveling for IBM, Guglielmi would typically bring along three administrative assistants. Working at Taligent, he now sometimes travels alone, picks up his own rental cars, and communes with his laptop in hotel rooms. His office at work is close to the engineering staff quarters. ''I'm 'in' the technology now,'' he says. ''I'm much closer to where the world is today.'' When the directors have finished their search and named their choice, the IBM establishment can likewise move much closer to where the world is today. Akers steered diligently in that direction, but with meager success for most of his years and none at the end. The next contender may find the battle equally difficult. But somewhere there is someone who can run this company and do it well. May he soon appear -- for the sake of the stockholders, the employees, and the might of the nation.

CHART: NOT AVAILABLE CREDIT: NO CREDIT CAPTION: A PATTERN OF DECLINE AT IBM REVENUES PROFITS EMPLOYEES MARKET CAPITALIZATION VALUE OF AKERS'S STOCK

CHART: NOT AVAILABLE CREDIT: SOURCE: MCKINSEY & CO. CAPTION: A PATTERN OF DECLINE AT IBM MARKET SHARE