Motorola's 1986 bet on near-zero defects made Six Sigma the world's quality standard

Through the late 1970s and early 1980s Motorola's quality was being beaten by Japanese electronics. The company ran a formal program to raise quality tenfold in five years and made training managers a top priority, but quality was still a campaign, not a number. Engineer Bill Smith argued the fix was statistical: engineer processes so six standard deviations fit between the process mean and the nearest failure point.

Measure quality as defects per million opportunities and aim every process at six sigma — 3.4 defects per million — with trained experts driving each project.

Smith had spent nearly 35 years in engineering and quality assurance, and he correlated field failures with defects caught inside the factory — showing that raising internal standards cut warranty costs downstream. CEO Bob Galvin, prodded by quality champion Art Sundry, turned the idea into a corporate goal. Mikel J. Harry built the training and deployment machinery, and Motorola put full-time quality specialists on improvement projects defined by the DMAIC cycle: define, measure, analyze, improve, control.

The goal was unambiguous: 3.4 defects per million opportunities per process. Motorola won the inaugural Malcolm Baldrige National Quality Award, and after Jack Welch made Six Sigma central at GE in 1995, the method spread through Johnson & Johnson, Texas Instruments and much of corporate America — it remains one of the most widely used quality-management systems.

In the late 1970s Motorola was losing the quality war to Japanese electronics makers. The company ran a formal program to raise quality tenfold in five years and treated management training as its most important asset, but quality was still a campaign with a start and an end. What changed in 1986 was the unit of account.

Engineer Bill Smith showed that failures in customers' hands correlated with defects caught inside the factory — so the way to cut warranty costs was to raise internal standards. Six Sigma reframed the goal in statistics: a process running at six sigma fits six standard deviations between its mean and the nearest specification limit, about 3.4 defects per million opportunities. CEO Bob Galvin, pushed by executive Art Sundry, made it a company-wide goal, and Mikel J. Harry built the training machinery that carried it.

The scoreboard was portable, and portability did the rest. Motorola won the inaugural Malcolm Baldrige National Quality Award; when Jack Welch made Six Sigma central at GE in 1995, adoption swept through Johnson & Johnson, Texas Instruments and most large American companies. Six Sigma remains one of the most widely used quality-management systems in the world.

It converted a vague aspiration — better quality — into a single measurable number that any process in any plant could be scored against.

It tied internal defect rates to field failures and warranty costs, so executives could see quality as money rather than virtue.

It created full-time quality specialists, making improvement a discipline rather than a temporary campaign.

The metric was statistical, so any company could adopt the same scoreboard — which is how the method outgrew Motorola.

A quality goal works when it is one unambiguous number that any process can be audited against, backed by specialists whose full-time job is to move it — not when it is a slogan on the wall.

Six Sigma outgrew its birthplace faster than almost any method before it: after Jack Welch made it central at GE in 1995, it spread through Johnson & Johnson, Texas Instruments and much of corporate America. The critiques came almost as fast. Consultant Philip Crosby argued in 1996 that 3.4 defects per million was inadequate for semiconductors, whose units pack more than a million components, and later writers noted a survivorship bias — Amazon and Google thrive while Sears and Computer Sciences Corporation, both Six Sigma shops, failed. Motorola itself split in two after years of losses.