Because it has become so consistently successful, I.B.M. is almost boring. This is a company so predictable that its financial forecast is packaged as a “five-year road map,” as if it were some sort of state planning exercise.

Yet behind I.B.M.’s relentless progress over the last decade is a game plan that has been anything but conservative. The company shed multibillion-dollar businesses. It chose higher profit margins over corporate size, and expanded aggressively overseas, seeking sales, low-cost engineering talent and quicker organizational reflexes.

Investors, however, haven’t been bored. The company’s stock price has surged. In November, Warren E. Buffett, who typically shuns technology stocks, announced he had accumulated $10 billion of I.B.M. shares, a stake of more than 5 percent.

All of that didn’t just happen. A large portion of the credit goes to Samuel J. Palmisano, who steps down on Sunday after nearly a decade as chief executive. During his tenure, I.B.M. has been a textbook case of how to drive change in a big company — when so much of the study of business innovation focuses on start-ups and entrepreneurs.

This column is a glimpse of the thinking behind some of the major steps I.B.M. has taken under Mr. Palmisano’s leadership, based on two recent interviews with him.

He says his guiding framework boils down to four questions:

• “Why would someone spend their money with you — so what is unique about you?”

• “Why would somebody work for you?”

• “Why would society allow you to operate in their defined geography — their country?”

• “And why would somebody invest their money with you?”

Mr. Palmisano formulated those questions in the months after he became C.E.O. in March 2002 His predecessor, Louis V. Gerstner Jr., recruited to I.B.M. in 1993, had already pulled the company out of a financial tailspin, first reducing the size of the work force and cutting costs, and then leading a remarkable recovery.

In meetings after he took over, Mr. Palmisano told colleagues that I.B.M. was still good, but that it wasn’t the standard-setting corporation that it had been when he joined in 1973. (A history major at Johns Hopkins and a star offensive lineman on the football team, he turned down a tryout with the Oakland Raiders of the N.F.L. for a sales job at the company.)

The four questions, he explains, were a way to focus thinking and prod the company beyond its comfort zone and to make I.B.M. pre-eminent again. He presented the four-question framework to the company’s top 300 managers at a meeting in early 2003 in Boca Raton, Fla.

“This needs to be our mission and goal, to make I.B.M. a great company,” he said, according to executives who attended the gathering.

THE pursuit of excellence in those four dimensions shaped the strategy. To focus on doing unique work, with its higher profits, meant getting out of low-margin businesses that were fading. I.B.M.’s long-range technology assessment in 2002 concluded that the personal computer business would no longer present much opportunity for innovation, at least not in the corporate market.

The hub of innovation would shift to services and software, often delivered over the Internet from data centers, connecting to all kinds of devices, including PCs. Today, that is called cloud computing; when I.B.M. started promoting the concept several years ago the company called it on-demand computing.

So Mr. Palmisano led a lengthy strategic review of the PC business, deciding to sell while it was still profitable. Internal arguments against a sell-off were intense: PCs pulled in sales of other I.B.M. products in corporate accounts, the cost of electronic parts for its larger computers would jump without the purchasing power of its big PC division, and the corporate brand and its reputation would suffer without PCs, the one I.B.M. product touched by millions of people.

Lately, Hewlett-Packard has engaged in a similar debate, first declaring that it was looking to sell its PC business, then backing off. “I’ve heard every one of the arguments, every one of them,” Mr. Palmisano says. “But if you decide you’re going to move to a different space, where there’s innovation and therefore you can do unique things and get some premium for that, the PC business wasn’t going to be it.”

In 2004, I.B.M. sold its PC business to Lenovo of China. Mr. Palmisano says he deflected overtures from Dell and private equity firms, preferring the sale to a company in China for strategic reasons: the Chinese government wants its corporations to expand globally, and by aiding that national goal, I.B.M. enhanced its stature in the lucrative Chinese market, where the government still steers business.

In total, the PC, disk drive and other hardware businesses that Mr. Palmisano sold off generated nearly $20 billion a year in sales, if not a lot of profits.

The divestitures meant that I.B.M. was no longer the world’s largest information technology company. Hewlett-Packard took that title and took a different strategic path as well, doubling its bet on PCs by acquiring Compaq in 2001. “You see the choice that was made, and how the economics worked out,” Mr. Palmisano observes.

Today, I.B.M.’s stock market value, at $217 billion, is more than four times that of the struggling H.P.

I.B.M. invested heavily elsewhere, buying the business consulting firm PricewaterhouseCoopers Consulting, for $3.5 billion in 2002, for its expertise in specific industries. For I.B.M., the emphasis was to move up from selling customers computers and software to helping them use technology to solve business challenges in marketing, procurement and manufacturing.

Corporations and governments are drowning in a flood of data from internal systems and the Web, struggling to make sense of it. To get ahead of that challenge, I.B.M. has spent more than $14 billion since 2005 buying 25 software companies that specialize in data mining and analytics, looking for useful patterns in data in fields as varied as disease treatment, traffic management and crime detection. And it has increased its research and development budget by 20 percent under Mr. Palmisano, to about $6 billion a year.

Combining research, specialized skills and sophisticated technology is the recipe behind I.B.M.’s Smarter Planet initiative, begun in 2008. It now has more than 2,000 projects worldwide, applying computer intelligence to create more efficient systems for utility grids, traffic management, food distribution, water conservation and health care.

The idea, Mr. Palmisano explains, is to “go to a space where you’re uniquely positioned and use the value of I.B.M.’s integration.”

For him, the Smarter Planet effort is a return to I.B.M.’s roots. Shortly before he became chief executive, he dipped into corporate archives, reading speeches and memos from the founder, Thomas Watson Sr. When Mr. Palmisano was an executive assistant to John F. Akers, then the chairman, in 1989 and 1990, he had lunch with Thomas J. Watson Jr., a former chairman, once a month. The Watsons, he says, always defined I.B.M. as a company that did more than sell computers; they believed that it had an important role to play in solving societal challenges.

“It’s old-fashioned, but it’s motivational,” he says.

And it resonates with the young people I.B.M. is recruiting these days, he says. A couple of times a year, Mr. Palmisano speaks to groups of elite students whom I.B.M. is trying to woo to its research labs. The pitch, he says, is that I.B.M. is a place where you can make a difference and do deep science.

“You can change the world, and you can compete for a Nobel prize,” he says, referring to I.B.M.’s five Nobel winners.

(Eighty-seven percent of the candidates who were offered jobs by I.B.M. Research this year joined the company.)

Mr. Palmisano’s appeal to young technologists is just one example of an answer to one of his four questions. All four, he says, must be addressed. That challenge is now passed to Virginia M. Rometty, who this week becomes the ninth chief executive in the company’s 100-year history.

“The hardest thing is answering those four questions,” Mr. Palmisano says. “You’ve got to answer all four and work at answering all four to really execute with excellence.”

Noah Kravitz tweeted for Phonedog as @Phonedog_Noah, but later changed his username when he left the company - taking 17,000 followers with him.

The company is now seeking damages of $2.50 (£1.60) per user, per month - a total of $370,000.

Mr Kravitz said his former employer had given him permission to continue using the account after he left.

He told the New York Times that Phonedog had allowed him to make the account personal as long as he agreed to "tweet on their behalf from time to time".

The 17,000 followers, which have since risen to 22,000, had been built up by Mr Kravitz during his four years at the company where he worked as a blogger.

However, eight months later the company filed a lawsuit claiming that the account's followers were a customer list, and that it had invested "substantial" resources into building it.

In a written statement, it said: "The costs and resources invested by Phonedog Media into growing its followers, fans and general brand awareness through social media are substantial and are considered property of Phonedog Media.

"We intend to aggressively protect our customer lists and confidential information, intellectual property, trademark and brands.""Companies will now be developing careful ways of deciding if they want to tweet with a conjoined account," said Barbara Cookson, an intellectual property lawyer in the UK.

"For ordinary businesses it's quite difficult to gain a following without a strong personality. You have to have a very strong brand for it to work."

Ms Cookson argued it is hard to pinpoint a financial value to Twitter followers as it is unclear why they follow a particular account.

It's arguable as to whether a Twitter follower list is comparable to a mailing list.

"If Phonedog has been using it to run offers, it perhaps is a mailing list that has value."

However, intellectual property solicitor Leigh Ellis said Phonedog are likely to have a strong case as the original account featured the company's name.

"Let me put it this way, I'd prefer to be on Phonedog's side," he told the BBC.

"If you're a follower, who are you following? You might be following Noah, but it's PhonedogNoah. There's a very good argument that the reputation accrued is to the company, rather than the individual."

Started by Eric Ries, a 32-year-old technology entrepreneur from San Francisco, it is a philosophy of how to start and run a business on almost nothing but fumes, hard work and a willingness to engage endlessly with your customers. Ries this month published his first book, Lean Start-Up, which has quickly caught America's entrepreneurial imagination.

The basic idea is that as technology costs have fallen and the potential for almost effortless online collaboration has risen, it is time to rethink how companies are born. Under the old model, entrepreneurs would scratch their heads, come up with an idea, raise money, spend it developing a product then unveil it - "ta-da!" - to the public, hoping it would catch on. There might then be a scramble to refine it in response to customer feedback or another fortune spent trying to advertise or market the product to success.

These days, Ries argues, customers are much more willing to engage in the development of products early on, and it is the duty of entrepreneurs to take advantage of that. For example, if you have a technology firm, you should release a "minimal viable product" as early as possible, see how people use it and respond and then improve it and repeat. From the outset, then, a start-up is testing, searching and probing the market at very low cost, only investing serious resources once it is confident of success.

For example, say you are developing a new music app. You could spend months perfecting it based on nothing but your vision of what the market requires. Or you could test a very early version of that vision and keep refining it in response to your consumers.

Entrepreneurs should not be afraid any more of releasing incomplete products as customers, especially enthusiasts in your field, are all too happy to participate in helping you improve it along the way. It has worked for tech successes like Dropbox, a fast-growing file-sharing company, and even Google, which releases "beta" versions of its products expecting them to be improved as they learn how users behave.

Ries's advice is now sought not just by start-ups but also major corporations eager to find ways to innovate more cheaply and more quickly. Even government, he says, could benefit from road-testing ideas on a small scale before rolling them out to huge markets at great expense. "How can we learn more quickly what works and discard what doesn't?" Ries asks in his book. It is a question everyone would like the answer to.

(c) Philip Delves Broughton

(c) Mark McGuinness

(C) Emma Haslet, Management Today