
Six Sigma, as this account defines it, is a company-wide effort to study defects statistically, remove their causes, and reach a level of 3.4 defects in a million items. The method, it says, began in the United States in the mid-to-late 1980s: Motorola introduced it in the mid-1980s, and GE’s results in the mid-1990s stirred other firms. About 20 percent of U.S. companies had adopted it, the account says, and in Korea, large firms such as Samsung and LG had set up dedicated teams. Those “current” remarks belong to the time of writing, whose sources include a 2002 Samsung Economic Research Institute book; they are not a 2026 industry census. Samsung SDI applied Six Sigma in production in 1996, after display-business profits had weakened, dropping an earlier quality method. Training began in 1997. In 1998 it reached office work as well. People were trained as Black Belts (a target of 5 percent of employees) and Green Belts (50 percent). Work ran in three forms: SQM to strengthen manufacturing, a quality-qualification system from executives to all staff, and C-S-I projects. SQM was used at new plants — electron guns, flat panels, Cheonan — with line checklists that included standard maintenance and audit, MAS (gauge R&R), and Cp/Cpk (SPC), plus a regular PDCA cycle. The qualification ladder is White Belt (study and a test), Green Belt (project and exam), Black Belt (project, exam, and a teaching role), and Champion (a management qualification, with results shared on Champion’s Day).
Projects follow a C-S-I cycle: Chart selects and defines a product-level problem by analyzing variation and defects; Solve lists process variables, ranks them, and optimizes; Implement puts the chosen remedies to work. Critical-to-quality characteristics are chosen first from the customer’s point of view. The account names a U.S. consultant, Dr. Bajaria of Multifacet, as the person who led the introduction. From 1997, cost savings of about 90 billion won were reported. From 2000, the last Saturday of each month was Champion’s Day, for sharing know-how and results. In 2001, some 2,000 projects were said to have brought 250 billion won in financial results. Management took the training itself, and belt holders were favored in promotion. Introduction, the account warns, is a long game: two to three years to prepare, two to three more to see results, with process standards and office IT in place first. A chief executive’s knowledge and passion matter; Jack Welch at GE is the example given. Statistical thinking — standard deviation, statistical quality control, design of experiments — should be the base. It should not become a passing fashion. Failures named are unconditional fantasy, treating it only as a problem-solving toolkit, passive leadership, and blind imitation. A fuller Korean report on the same ground is on CampusDown.
Six Sigma, as this account defines it, is a company-wide effort to study defects statistically, remove their causes, and reach a level of 3.4 defects in a million items. The method, it says, began in the United States in the mid-to-late 1980s: Motorola introduced it in the mid-1980s, and GE’s results in the mid-1990s stirred other firms. About 20 percent of U.S. companies had adopted it, the account says, and in Korea, large firms such as Samsung and LG had set up dedicated teams. Those “current” remarks belong to the time of writing, whose sources include a 2002 Samsung Economic Research Institute book; they are not a 2026 industry census.
Samsung SDI applied Six Sigma in production in 1996, after display-business profits had weakened, dropping an earlier quality method. Training began in 1997. In 1998 it reached office work as well. People were trained as Black Belts (a target of 5 percent of employees) and Green Belts (50 percent). Work ran in three forms: SQM to strengthen manufacturing, a quality-qualification system from executives to all staff, and C-S-I projects. SQM was used at new plants — electron guns, flat panels, Cheonan — with line checklists that included standard maintenance and audit, MAS (gauge R&R), and Cp/Cpk (SPC), plus a regular PDCA cycle.
The qualification ladder is White Belt (study and a test), Green Belt (project and exam), Black Belt (project, exam, and a teaching role), and Champion (a management qualification, with results shared on Champion’s Day). Projects follow a C-S-I cycle: Chart selects and defines a product-level problem by analyzing variation and defects; Solve lists process variables, ranks them, and optimizes; Implement puts the chosen remedies to work. Critical-to-quality characteristics are chosen first from the customer’s point of view. The account names a U.S. consultant, Dr. Bajaria of Multifacet, as the person who led the introduction. From 1997, cost savings of about 90 billion won were reported. From 2000, the last Saturday of each month was Champion’s Day. In 2001, some 2,000 projects were said to have brought 250 billion won in financial results. Those sums are left in their years.
Management took the training itself, and belt holders were favored in promotion. Introduction, the account warns, is a long game: two to three years to prepare, two to three more to see results, with process standards and office IT in place first. A chief executive’s knowledge and passion matter; Jack Welch at GE is the example given. Statistical thinking should be the base, and the method should not become a passing fashion. Failures named are unconditional fantasy, treating it only as a problem-solving toolkit, passive leadership, and blind imitation. A fuller Korean report on the same ground is on CampusDown.
Last updated: 14 November 2026
3.4 defects in a million items, by studying defects statistically and removing their causes, as a company-wide effort.
Production in 1996, training in 1997, offices in 1998, Champion’s Day from 2000. Savings of about 90 billion won from 1997, and 250 billion won from some 2,000 projects in 2001, are left in those years.
Chart selects and defines a product-level problem; Solve lists and ranks process variables and optimizes them; Implement puts the chosen remedies to work. Critical-to-quality features are chosen first from the customer’s point of view.
Refund policy