Do It Right the First Time!

"There's no reason to have a Plan B because it distracts from Plan A." - Will Smith

After WWII decimated most of the European and Asian manufacturing facilities, the US companies returned to manufacturing goods needed by much of the rest of the world, without concern for quality. Products were needed, now. American automobiles were being built and sold, primarily with only styling changes. Innovation and quality were quashed, leaving such companies as Tucker and Studebaker in their wake. Remember the Corvair (unsafe at any speed), the Pinto with gas tank fires and Firestone 500 tires.

Meanwhile, Japan was being rebuilt under the Marshal Plan. Known for cheap manufacturing and goods, Japan knew that it’s innovation and performance built into its war machine could be put into its consumer goods. In 1950, W.E. Deming spoke at the Hakone Convention Center in Tokyo with a speech on “Statistical Product Quality Administration". Put into practice, it showed that quality is an outcome of the process.

In 1951, the Japanese Union of Scientists and engineers (JUSE) established the Deming Prize as Japan’s national quality award for industry and services. Still, Japan was perceived as the source of cheap, low quality goods.

Then in 1973 came the Arab oil embargo. Americans now wanted fuel efficient cars and Europe and Japan, primarily Japan were the best source for them. The result was that Americans discovered the quality of Japanese, German, and Swedish cars. American companies started losing their market share. The American companies complained about cheap Japanese labor but quality was the real culprit.

In 1980, NBC broadcasted the show, “If Japan Can… Why Can’t We” that introduced America to W.E. Deming and Statistical Quality Control. Ford hired Deming in 1981 and was quickly surprised that Dr. Deming concentrated on Ford’s management, claiming that 85% of Ford’s quality problems were directly attributable to management. As a matter of fact, Dr. Deming was opposed to performance reviews for the same reason. The employees need the proper training, and tools. The services and manufacturing needed the correct process. And so on.

In the processes of engineering and assembling automobiles, the practice of mass inspection was executed at the end of the assembly line, if something didn't work or fit or look right, the vehicle was pulled aside and fixed. Mind you, the engineering, parts, and assembly costs were already most of the cost of the vehicle. Fixing it added to the cost and reduced the profit. If part of the manufacturing, a body panel stamping for example, was outside of the specifications and not discovered until the first fault was found at the end of assembly, how many additional vehicles have the same defect, especially with another vehicle reaching the end every 20 to 30 seconds?[1]

Computers and their software applications also underwent a transformation for quality. As applications became more complex and teams of developers increased in size, processes and safeguards needed to be established and enforced. The SDLC, Software Development Life Cycle was defined and refined and updated for the unique environments of different companies. Also, Source Code Control Systems (SCCS) were created. Those tools were designed to ensure that no more than one developer could work on a code file at the same time.

Unfortunately, there is no automated check of the code to test for simple mistakes and errors that frequently get into production and may impact data security which impacts customer satisfaction and confidence. Like the final inspection on an assembly line, the software industry has created a service model for handling these errors.

In the SDLC, there are inspections that are required, called code reviews or peer reviews, that step through the code, line by line, by peers of the developer, so that problems like the scope of variables in a program or loading of variables from a database, will not allow a variable to carry over user data from a previous call to the database, thus exposing one customers information to another customer. That real-life error was caused by the way that the variable was defined, by missing data from the database, and by how the program processed the occurrence of missing data.

There is a discipline, the Software Engineering Institute Capability Maturity Model Integration, SEI/CMMI, that if properly followed, would catch one of the errors above, if not all three. Thankfully, the U.S. Air Force does require the use of the highest level of CMMI for its software.

Projects and Sales usually fail because they do not address the people issues involved. People are the source of the best information, the greatest innovation, and the best designs. People provide commitment, dedication, loyalty, communication, and resolutions. People define the products that they want, they buy the products, and they use the products. But people also fail, resist, withhold information, have personal lives, get sick, and have needs and desires. Those are issues that one of Dr. Deming's 14 principles addresses. The transformation of an organization is everybody's job.

Back to Plan B vs. Plan A. Above are two instances where Plan A was not fully defined and implemented, requiring a Plan B to fix the errors. Plan B increased the overall cost of the products through the "Fix It" process, where the correct design, planning, and execution of Plan A may have had only a minor increase in costs and actually could have been applied to other, similar products, reducing overall engineering and production costs for the corporation as long ss they follow Dr. Deming's point 9, "Break down barriers between departments".

[1] W. Edwards Deming, "Out of the Crisis", pp. 28-31
[2] Philip B. Crosby, "Quality is Free"