In the spring of 1993, IBM was the most famous computer company in the world, and it was coming apart. It had recorded the first annual loss in its history in 1991 and a larger one in 1992. In 1993 it would lose more than $8 billion. Its mainframe business, the profit engine for decades, was being undercut by smaller and cheaper machines, and its market value had fallen to about $29 billion. The plan already in motion under the previous CEO, John Akers, was to split IBM into a set of more independent businesses that could each go their own way.
The board went outside for help. Lou Gerstner had never run a technology company. He had been a McKinsey consultant, a senior executive at American Express, and most recently chairman and CEO of RJR Nabisco. He took over IBM in April 1993, and plenty of people in the industry doubted that a food-and-tobacco executive could understand the computer business. Gerstner did not pretend to know what he did not know. He set out to find out.
Within weeks, he launched an effort that became known as Operation Bear Hug. Roughly 250 of IBM's senior managers were each told to visit at least five of the company's major customers within three months. These were not sales calls. The assignment was to listen to what customers were struggling with, think about how IBM could help, and report back. Gerstner spent much of his own early months in the same way, including a meeting with the chief information officers of large customers at an IBM center in Chantilly, Virginia.
What came out of that listening ran straight into the breakup plan. Big customers did not want to buy a pile of separate parts from a pile of separate vendors and then wire everything together themselves. They needed someone to take responsibility for making the whole system work. IBM was one of very few companies that made nearly every piece, from chips and software to mainframes, storage, and PCs, and it had people around the world to support them. Gerstner concluded that cutting IBM into pieces would destroy the one advantage customers valued most.
So he stopped the breakup. It was the most important decision of his tenure, and it cut against the conventional wisdom of the day. Instead of carving IBM up, he set out to make it work as one company organized around solving customers' problems. Services, which meant helping customers plan, build, and run their technology, grew from a side business into a central one. He was blunt about priorities, too. That summer he told reporters that “the last thing IBM needs right now is a vision.” The first job was to stop the bleeding and execute.
The fixes were painful. IBM had long been known for something close to lifetime employment, and that ended. Over the course of the turnaround the company cut more than 100,000 jobs. Gerstner also went after a culture that had grown inward-looking, slow, and proud. Early on he told Fortune that it was “time for IBM to perform and then talk, instead of talk and then perform.” He pushed managers to measure success by how customers fared, not by how their own division did against the division down the hall.
The results came steadily rather than overnight. By the time Gerstner stepped down as CEO in March 2002, IBM's market value had climbed to about $168 billion, nearly six times what it was when he arrived. Services had become a large and reliable source of revenue, and IBM had reinvented itself as the company large organizations called when they needed technology to work as a whole. A company that many experts had wanted to dismantle became the example people studied for how to save a giant.
The story has its costs, and they are worth naming. Tens of thousands of people who had expected to retire from IBM lost their jobs. Some product lines Gerstner inherited were later abandoned, and two years after he left, IBM sold its PC business to Lenovo. No turnaround of that size is clean. Gerstner died in December 2025, and the tributes that followed returned again and again to the same point: he kept the company together because he paid attention to what customers would need next.
Notice what Gerstner did not do. He did not sit in a conference room and choose between the plans that smart insiders had already drawn up. The breakup plan had been built by capable people looking at real numbers. What it lacked was the most important information of all, which was sitting in customers' offices waiting for someone to ask. Pilots are required to become familiar with all available information before a flight. Gerstner treated his first months at IBM the same way.
You can do this on Monday. Pick five people your team serves, whether they are outside customers, another department, or the crew on the next shift. Go see them in person. Do not pitch, defend, or explain. Ask what is getting in their way and what they wish your team did differently, then write down exactly what they say. Bring it back and lay it next to your current plan. If the two do not match, believe the customers before you believe the plan.
The answer to your hardest problem may already be sitting with the people you serve. They will usually tell you, if you go and ask and stay quiet long enough to hear it. The leaders who skip that step are not short on intelligence. They are short on information, and no amount of cleverness in the conference room can make up for what they never went out to learn.