That Day the CPK Report Came Out, and the Room Fell Silent for Three Seconds
I still remember years ago, our factory just took on a project for a new client. Everyone was in high spirits then, feeling like we were finally going to get rid of those trivial demands from old clients. However, when the CPK report for the first trial production run came out, the main process parameter CPK was only 1.08, and the DPMO was as high as 6210. The entire meeting room instantly fell silent; even the sound of the air conditioning was crystal clear. I was sitting in the corner then, watching my supervisor's face turn pale, and I thought to myself: "It's over. Not only will we be working overtime, but we'll practically be living at the factory." Honestly, everyone's attitude towards quality was "good enough is fine."
Where's the Problem? "Good Enough" Is the Biggest Problem
You know what? Many times, we talk about quality, but what we really think is "just don't screw up." This "good enough" culture is actually the biggest killer of quality. To put it plainly, everyone felt that specifications and standards were just for QA, and as long as the final product passed, whatever happened in between didn't matter. This is like building a house where you only care about the appearance, ignoring an unstable foundation. Only when something truly goes wrong do you realize how serious the consequences of "good enough" are. That's how we were then; everyone just focused on completing their work orders, but no one really thought about why this parameter's CPK was so low? What were the underlying reasons?
What to Do in Practice? Start Small to Build "Quality Awareness"
To change this "good enough" culture, frankly, is not something that can be achieved overnight. After that time we were 'fried' by the client, we reflected deeply and started working on several areas.
- Data Transparency to Highlight Problems: Previously, CPK reports were only shared with supervisors. Now, we post the CPK and DPMO for all process parameters on the production line bulletin boards, and even clearly mark the weekly yield target achievement rates. This allows everyone to see how their responsible area is performing. When the data is laid bare in front of you, it's hard to pretend you haven't seen it.
- Establish an "Anomaly Shutdown" Mechanism: Before, if a process parameter drifted slightly but was still within specification, everyone would continue production. Later, we introduced an "anomaly shutdown" mechanism: if a critical parameter's CPK falls below 1.33, production must stop immediately for review, root cause identification, and resolution. Initially, there were many complaints, feeling it impacted production capacity, but over time, everyone proactively started monitoring their own processes because no one wanted to be the one to halt the production line.
- Training Is More Than Just "Lectures": We started bringing training to the field, operating machines directly, simulating various abnormal conditions, allowing engineers and technicians to personally experience the consequences of parameter deviation. This was far more effective than listening to lectures in a meeting room.
So the key is to make everyone feel a connection to quality data, to let them know that these numbers are not just QA's concern but an integral part of their daily work.
The Most Common Pitfall: Supervisors' Double Standards of "Only Looking at Output, Not Quality"
When we were promoting a quality culture, the most common pitfall we encountered was the "double standards" of certain supervisors. They would talk about quality, but as soon as production pressure arose, they would start turning a blind eye. I remember one time, an engineer proactively stopped production for review due to process parameter deviation, only to be scolded by the supervisor for "wasting production capacity." After this incident spread, everyone started reverting to the "good enough" mode. Later, our senior management directly intervened, explicitly stating: "Quality is the first priority; output can be chased, but quality absolutely cannot be compromised." Only by incorporating quality metrics into supervisors' performance reviews did this trend slowly get corrected. Honestly, if supervisors don't value quality, no matter how hard your subordinates work, it will be in vain.
One Thing You Can Do Today
Post your critical process data where you can see it.
Article Category: Quality Assurance Management