Work-Outs, however, weren’t perfect. They took an entire day of everyone’s time and usually meant the plant had to slow down production so that workers could all attend the meetings. It was something a division or plant could do once or twice a year, at most. And though the Work-Outs left everyone feeling excited and hungry for change, the effects were frequently short-lived. A week later, everyone was back at their old jobs and, often, their old ways of thinking.
Kerr and his colleagues wanted to foster perpetual ambitions. How, they wondered, do you get people to think expansively all the time?
IV.
In 1993, twelve years after becoming chief executive of General Electric, Jack Welch traveled to Tokyo and, while touring a factory that made medical testing equipment, heard a story about Japan’s railway system.
In the 1950s, during the lingering wake of the devastation of the Second World War, Japan was intensely focused on growing the nation’s economy. A large portion of the country’s population lived in or between the cities of Tokyo and Osaka, which were separated by just 320 miles of train track. Every day, tens of thousands of people traveled between the cities. Vast amounts of raw industrial materials were transported on those rail lines. But the Japanese topography was so mountainous and the railway system so outdated that the trip could take as long as twenty hours. So, in 1955, the head of the Japanese railway system issued a challenge to the nation’s finest engineers: invent a faster train.
Six months later, a team unveiled a prototype locomotive capable of going 65 miles per hour—a speed that, at the time, made it among the fastest passenger trains in the world. Not good enough, the head of the railway system said. He wanted 120 miles per hour.
The engineers explained that was not realistic. At those speeds, if a train turned too sharply, the centrifugal force would derail the cars. Seventy miles an hour was more realistic—perhaps 75. Any faster and the trains would crash.
Why do the trains need to turn? the railway head asked.
There were numerous mountains between the cities, the engineers replied.
Why not make tunnels, then?
The labor required to tunnel through that much territory could equal the cost of rebuilding Tokyo after World War II.
Three months later, the engineers unveiled an engine capable of going 75 miles per hour. The railway chief lambasted the designs. Seventy-five miles per hour, he said, had no chance of transforming the nation. Incremental improvements would only yield incremental economic growth. The only way to overhaul the nation’s transportation system was to rebuild every aspect of how trains functioned.
Over the next two years, the engineers experimented: They designed train cars that each had their own motors. They rebuilt gears so they meshed with less friction. They discovered that their new cars were too heavy for Japan’s existing tracks, and so they reinforced the rails, which had the added bonus of increasing stability, which added another half mile per hour to cars’ speed. There were hundreds of innovations, large and small, that each made the trains a little bit faster than before.
In 1964, the Tōkaidō Shinkansen, the world’s first bullet train, left Tokyo along continuously welded rails that passed through tunnels cut into Japan’s mountains. It completed its inaugural trip in three hours and fifty-eight minutes, at an average speed of 120 miles per hour. Hundreds of spectators had waited overnight to see the train arrive in Osaka. Soon other bullet trains were running to other Japanese cities, helping fuel a dizzying economic expansion. The development of the bullet train, according to a 2014 study, was critical in spurring Japan’s growth well into the 1980s. And within a decade of that innovation, the technologies developed in Japan had given birth to high-speed rail projects in France, Germany, and Australia, and had revolutionized industrial design around the world.
For Jack Welch, this story was a revelation. What GE needed, he told Kerr when he got home from Japan, was a similar outlook, an institutional commitment to audacious goals. Going forward, every executive and department, in addition to delivering specific and achievable and timely objectives, would also have to identify a stretch goal—an aim so ambitious that managers couldn’t describe, at least initially, how they would achieve it. Everyone, Welch said, had to partake in “bullet train thinking.”
In a 1993 letter to shareholders, the chief executive explained that “stretch is a concept that would have produced smirks, if not laughter, in the GE of three or four years ago, because it essentially means using dreams to set business targets—with no real idea of how to get there. If you do know how to get there—it’s not a stretch target.”
Six months after Welch’s trip to Japan, every division at GE had a stretch goal. The division manufacturing airplane engines, for instance, announced they would reduce the number of defects in finished engines by 25 percent. To be honest, the division’s managers figured they could hit that target pretty easily. Almost all the defects they found on engines were small, cosmetic issues, such as a slightly misaligned cable or unimportant scratches. Anything more serious was corrected before the engine was shipped. If they hired more quality assurance employees, managers figured, they could reduce cosmetic defects with little effort.
Welch agreed that reducing defects was a wise goal.
Then he told them to cut errors by 70 percent.
That’s ridiculous, managers said. Manufacturing engines was such a complicated affair—each one weighed five tons and had more than ten thousand parts—that there was no way they could achieve a 70 percent reduction.
They had three years, Welch said.
The division’s managers started panicking—and then began analyzing every error that had been recorded in the previous twelve months. Simply hiring more quality assurance workers, they quickly realized, wouldn’t do the trick. The only way to reduce errors by 70 percent was to make every single employee, in effect, a quality assurance auditor. Everyone had to take responsibility for catching mistakes. But most factory workers didn’t know enough about the engines to identify every small defect as it occurred. The only solution, managers decided, was a massive retraining effort.
Except that didn’t really work, either. Even after nine months of retraining, the error rate had fallen by only 50 percent. So managers started hiring workers with more technical backgrounds, the kind of people who knew what an engine ought to look like and, therefore, could more easily spot what was amiss. The GE factory manufacturing CF6 engines in Durham, North Carolina, determined that the best way to find the right employees was to hire only candidates with FAA certification in engine manufacturing. Such workers, however, were already in high demand at other plants. So to attract them, managers said employees could have more autonomy. They could schedule their own shifts and organize teams however they wanted. That required the plant to do away with centralized scheduling. Teams had to self-organize and figure out their own workflow.