Get Around the Box
by Joe Conklin
Thinking outside the box can be crucial to business success. But career success sometimes de-pends on getting around the box—and here I’m referring to the boxes on the organization chart.
When appealing up the chain of command is not practical, opportunities to address quality issues through lateral relationships might still exist. My introduction to this principle occurred early in my career at a company making electronic controls. I was in the quality assurance (QA) engineering branch.
Manufacturing was divided be-tween two groups. One produced the circuit boards, and the other packaged them into final units for testing and shipment. The product inspectors reported to the quality control branch. They documented any findings about a specific unit on the paperwork that followed it.
The inspectors also marked up a check sheet with data on the various types of defects they found. The check sheet summarized what took place during a shift and was eventually consolidated to understand trends at the level of an individual manufacturing station.
I was charged with taking the consolidation to the next level of generality—the entire factory—and reporting the overall yield and major defects weekly.
I presented my report during a meeting attended by representatives from all the major departments, such as manufacturing, production control, engineering, quality and purchasing.
Groups within the manufacturing department had differing opinions about the factory level yield report. The printed circuit board group accepted its share of the overall data without much comment and used it to shape its quality improvement efforts.
The representatives of the group assembling the boards were concerned the data made them appear responsible for things that were not their fault. They asked the quality group to add disclaimers and explanations to the report it would present at the departmental meeting.
On one occasion, representatives of the assembly group pointed out that a particular issue would soon not count as a defect because of an upcoming engineering change. But corporate engineering refused to change the drawing, claiming it would be too expensive. The assemblers argued they should not be held accountable when they were forced to use incorrect drawings.
How Much Justification?
The QA engineering manager, my immediate boss, explained to the assembly group’s leadership that our allocation of time in the weekly meeting was not likely to be increased, so there was not enough opportunity in that forum to go into the detailed level of justification it was requesting.
More important, the place where a defect was noted did not necessarily indicate the people working there were responsible. The cause could just as easily have occurred farther up the production stream.
My group’s goal was to merely provide some high level indication of where to focus problem solving efforts, not to affix blame.
The manufacturing assembly group did not accept this explanation. Although it could not stop the weekly compilation and reporting of the inspectors’ data, it stopped cooperating with the QA group as far as defect reduction efforts were concerned.
Duplication of Work
The assembly group started compiling its own data and adding up the defects noted on the production paperwork, not realizing it was duplicating work already being done by the inspectors. I found myself in an awkward position. My immediate boss made it clear he wanted me to continue what I was doing. After his discussions with manufacturing failed, action from higher up was needed to resolve the dispute. For whatever reason, no such action was taken, or at least I could find no sign of it.
One consequence of the manufacturing group’s decision to stop providing test data was a large reduction in one of the inputs to the weekly report: the results of electronic testing in the assembly area. The people performing the tests worked for assembly. When QA engineering personnel began compiling data at the factory level, we provided a form for testers to use to summarize their results. With the end of cooperation, testers rarely completed this form.
The lack of action up the chain of command left me unsure of what to do next. Fortunately, the lack of test data from assembly did not prove fatal.
The cooperation from the circuit board group, combined with the information provided by the inspectors working in the assembly group, provided sufficient clues that ultimately raised the overall factory yield by about five percentage points.
With a little more thought, perhaps I would have pursued a solution to the test data issue from a different direction. One eventually came at the initiative of the test engineering department.
A test engineer heard about the concept of analyzing yield at the factory level and consulted some of our data for history on certain problem units. When several of the forms were missing, I explained the difficulties in getting the forms from the assembly test area. Because his area was on better terms with the assembly group than mine was, he went to its management and proposed a log system for communicating its results to his department.
After this log system was in place, the test engineer arranged to pass along these results to me so they could be included in time for the weekly overall yield report. This filled in the last remaining piece of information and added some validity to our trend analysis.
Up to that point in my career, the habit of waiting for all the issues to be resolved up the chain of command was so ingrained in me that this simple idea of obtaining support from another department with a common need had not occurred to me.
In an ideal world, upper management might have insisted that the assembly test area send me the data directly, but sometimes the ideal thing does not happen.
In my nearly 20 years in quality since that experience, I have encountered similar roadblocks. When the chain of command is a little slow or unable to respond, I continue to ask myself whether some other part of the organization with the same need could cooperate on a more lateral solution to the problem.
Keep Your Supervisor Informed
I always make it a point to keep my immediate supervisor informed before trying this strategy. I can’t recall ever being turned down. He or she normally shares my frustration at the delay.
A side benefit of this strategy is what you can learn about the perspectives and challenges of other parts of the organization.
Admittedly, the results of this strategy are not always optimal, but they can be good enough to stimulate upper management to provide the support needed to deal with a problem.
This strategy is an example of the general principle of coming up with more than one way to solve a problem. Following this principle helps to separate the good quality professionals from the great ones.
JOSEPH D. CONKLIN is a mathematical statistician at the U.S. Department of Energy in Washington, DC. He earned a master’s degree in statistics from Virginia Tech and is a senior member of ASQ. Conklin is also an ASQ certified quality manager, quality engineer, quality auditor and reliability engineer.