Kaizen is the Japanese word for "improvement." It's become synonymous with a philosophy made famous by Japanese manufacturers, notably Toyota. "Kaizen" is supposedly the key to Japanese success in industries from cars to game consoles.

"Kaizen" involves improving processes. Instead of focusing on the product, the idea is to improve the way it's made, and those are two different things. Formally, it's using statistical methods to examine the variations in outcomes, and work out ways to get closer to the result you want. So, for example, instead of just saying that a factory has to produce pipes with tolerances of 1/1000th of an inch, managers would look at what the variation between parts is and try to see if that could be narrowed.

Practically, it means that in some Japanese factories every line worker and manager could stop production, if necessary to identify a problem, and then to work out a plan of action, and then to implement it and check to see that it works. That's often called the PDCA Cycle — Plan, Do Check, Act. This might sound obvious, but it was relatively new in the 1950s when management theory was just starting to get going.

The idea, though, didn't originate in Japan — not entirely, anyway. Just after World War II the occupation authorities sponsored American managers to move to Japan and oversee the rebuilding of the economy and to get the country moving again as fast as possible. (For related reading, see: What Was the Marshall Plan?)

One of those managers was William Edwards Deming. Deming isn't well-known outside management theory circles, but in the 1930s he invented the statistical sampling methods that the U.S. Census first used in 1940 and still uses today. After World War II he had a job under General Douglas MacArthur, consulting with the Japanese government about its own census. It was in 1950 that he started working with local Japanese manufacturers who were all interested in how to improve quality and reduce costs. Deming's position was that better quality in itself reduces costs and boosts productivity, which in turn improves market share. This is called statistical process control, and the core of the idea wasn't even original with Deming — it was first outlined in the 1920s, when the very first control charts, the ancestors of the computer flow chart, were invented. (For related reading, see: Understanding Japanese Keiretsu.)

Among the principles of a "Deming company" — as outlined by Rafael Aguayo, author of Dr. Deming: The Man Who Taught The Japanese About Quality — were some that were and are anathema to many modern managers. Punishing workers who fail to reach certain goals can actually be counterproductive, because employees who fear for their jobs will be more reluctant to offer suggestions. Errors in production, Deming said, are sometimes because of a problem in the manufacturing system itself. (For more, see: Vertical Integration.)

For example, buying from the lowest-cost supplier of raw materials might introduce problems that even the best worker can't solve or get around. If the steel used in making a part of a machine is of low quality, the part will be also.

Another principle is that while processes can be improved, they can never be fully optimized; perfect manufacturing doesn't exist. In addition, cutting costs, by itself, isn't going to make products any better. Focus on quality, on the other hand, will by itself reduce costs because there won't be as many mistakes to fix. (For related reading, see: What are Economies of Scale?)

Japanese companies adopted Deming's principles enthusiastically. And the rest is, as they say, history. The improvements in manufacturing were often incremental, but they added up.

American factories weren't ignorant of this principle, as it was common during World War II — military hardware has to work and work well. But the use of statistics to track problems fell by the wayside after the war, as many industries were too busy trying to meet the seemingly insatiable demand for American products. Making more products became more important than making better ones.

Companies such as Toyota Motor Corp. (TM), meanwhile, saw their future in producing high-quality cars that would create loyal customers. The company developed a reputation for quality that wasn't dented until a spate of recalls between 2009 and 2011, when some drivers reported unexpected accelerations. During the 1960s and 1970s Japanese electronics makers used the same principles to establish dominance on the world market that wasn't upended until the late 1990s. (For related reading, see: Recall Recoil: Will Toyota Recover?)

Ford Motor Co. (F), on the other hand, developed a reputation for poor quality vehicles by the late 1970s and early 1980s, the Pinto being the best-known example. It took years for that to change — and it was largely because Ford started consulting with Deming in the late 1980s as it developed the Taurus. (For related reading, see: 5 of the Car Largest Recalls in History.)

While the success of Japanese companies such as Toyota and Sony Corp. (SNE) was driven by the focus on processes, analysts such as author Brooke Crothers have said Kaizen can actually stifle innovation if too-lavishly followed; Crothers noted in 2012 that Sony seemed ill-equipped to respond to the rise of Apple Inc.'s (AAPL) iPod, and focused on incremental improvements to portable CD players. Japanese electronics manufacturers lost their dominance, and in some areas may never get it back.

Toyota never gave up the concept of Kaizen, and management maintained that the recalls happened when the company drifted away from the idea. Analysts, such as Jeffrey Liker and Tim Ogden, who co-authored Toyota Under Fire: Lessons for Turning Crisis Into Opportunity, opined that the problem of unintended acceleration wasn't a clear cut case of manufacturing error at all. Others noted that it was a design issue, rather than something that could be corrected in a manufacturing process. (For related reading, see: The Cost of an Auto Recall.)

Given that Kaizen is designed to improve processes, in a manufacturing environment, it doesn't follow that it lends itself to every business. It's less clear, for instance, that the methods Deming pioneered could be applied to flight attendants or insurance companies — though certainly, many books and articles have been written that try to apply its principles to a wide variety of industries. Even so, the fundamental ideas of Kaizen have been adopted in the U.S. and elsewhere, with allowances made for local culture. (For related reading, see: Supply Chain Management.)

The Bottom Line

Kaizen is a statistically-influenced way of doing things with an eye toward constant improvement. It has a track record of success, but it must be applied wisely. If followed too closely it can actually stifle innovation. (For more, see: Total Quality Management and Six Sigma.)