The "Firefighting Dilemma" of the Quality Manager: Why the Busier, the Messier?
Every morning, there are always more than 20 customer complaint emails in the inbox of Li Yang, the quality manager of a manufacturing enterprise. "The pitch deviation of this batch of gears is 0.3mm, and they jam during assembly." "The drain valves of the previous batch of washing machines are leaking, and customers are demanding a full refund." Just as he dialed the phone of the production line supervisor, news came from the workshop: "The 50 air conditioners that were reworked yesterday have been stopped again today due to the same heat exchanger leakage." Li Yang smiled bitterly and said, "I'm not a quality manager now. I'm a 'firefighter' - I put out the fire wherever it breaks out, but I can never plug the leaks."
Even more absurdly, in order to "improve quality", the company expanded its inspection team from 50 to 150 people, accounting for 15% of the production line employees. However, the customer complaint rate only dropped from 8% to 6%. "Inspection is like closing the stable door after the horse has bolted. It can only pick out defective products, but cannot solve the problem of 'why defective products are produced'," Li Yang gave an example. "For instance, the pitch deviation of gears is due to the fact that thermal expansion and contraction were not considered during mold design. Inspectors can pick out the ones with excessive deviation, but they cannot make the mold more accurate." This "post - event remedy" model is like using a bucket to catch the rain in a leaking house - if the leaky points are not repaired, it's useless to have a bigger bucket.
"Perceptual misalignment" of ISO9001: The bottom line is never the endpoint
Many Chinese enterprises regard the ISO9001 certificate as a quality pass and even a key to win customers. I've seen the most extreme case: in order to obtain the certificate, an enterprise asked a consulting company to copy a template to compile documents, and made up three months' operation records one week before the audit. The workshop was temporarily cleaned spotlessly. After getting the certificate, the documents were locked in the cabinet of the finance office, and the daily production still relied on the experience of old workers — the drawing indicates a tolerance of ±0.1mm, but the workers say ‘it's okay as long as it's about right’ because anyway the inspectors will pick out the substandard products.
However, in excellent multinational enterprises, ISO9001 is the "quality baseline". The quality director of a Fortune Global 500 company said, "If a company fails to meet the ISO standards, it means that it doesn't even have the most basic awareness of processes. Talking about quality is just empty talk." Their internal processes are stricter than ISO requirements: For document control, not only signatures are required, but also records of version modifications should be kept, and audits are conducted quarterly. When dealing with non-conforming products, it's not just about "reworking/scrapping". Instead, a "Root Cause Analysis Report" needs to be filled out - is it a design issue? A material issue? Or a process issue? And the implementation of improvement measures should be tracked. For them, ISO is the "admission ticket", not the "terminal station".
The "underlying logic" of American quality theory: Upgrading from a point to a system
In the 20th century, the "Five Great Masters" in the quality field in the United States essentially established a quality framework "from points to the system":
Shewhart's SPC (Statistical Process Control): Convert "post-event inspection" to "process monitoring" - use control charts to track injection molding temperature and welding current. If the data exceeds the threshold, stop the machine immediately to avoid batch defective products.
Deming's PDCA cycle: Form a closed-loop of "Plan-Do-Check-Act" —— A certain automobile enterprise uses PDCA to optimize the welding process. The current is adjusted from 100A to 120A, and the welding strength is increased by 20%.
Juran's "Quality Trilogy": Elevating quality from a "departmental matter" to a "corporate strategy" —— A home appliance enterprise has set the goal of reducing the "DPU (defects per unit) from 1.2 to 0.5" for the year, and the entire company is working on improvements around this goal.
Crosby's "Zero Defects": Transform "good enough" into "doing it right the first time" —— A certain aviation enterprise requires that "the assembly error of parts ≤ 0.01mm". Employees check their tools three times before operation because "the cost of rework is five times that of doing it right the first time".
The combination of these theories enables American enterprises to shift from "addressing individual problems" to "building systematic capabilities". For example, a mechanical enterprise uses SPC to monitor the accuracy of machine tools and PDCA to improve the heat treatment process. As a result, it has reduced defective products by 35% and costs by 20% in one year.
Design quality: Grasp the "source" of quality
There is a well - worn saying in the industry that few enterprises can actually achieve: the root of quality lies in design. According to statistics from an authoritative organization, 70% of product defects can be traced back to design. For example, the charging port of a mobile phone is prone to breakage. It's not that the workers failed to weld it properly, but that the number of plug - ins was not considered during the design process (the industry standard is 5000 times, while some enterprises only design it to withstand 2000 times). Another example is that the refrigerator door doesn't close tightly. It's not that the sealing strip was not glued properly, but that the hinge angle has a 1 - degree deviation.
A Fortune 500 home appliance company I serve has engraved "design quality" into its DNA:
1. "Translate" customer requirements into measurable design indicators
Use QFD (Quality Function Deployment) to convert fuzzy requirements into specific goals. For example, when a customer says "the refrigerator should keep food fresh", it can be broken down into "the weight loss rate of vegetables within 7 days ≤ 5%", "the total number of bacteria in meat within 3 days ≤ 100 CFU/g", and "the humidity is 85% - 90%". These goals are not randomly set but obtained by testing the weight loss of vegetables under different humidity conditions in 1000 families.
2. Use DFMEA to "step on the brakes" in advance
Conduct DFMEA (Design Failure Mode Analysis) for each design project. For example, when designing the fixed structure of an air - conditioning compressor, list failure modes such as "screw loosening" and "bracket fracture". Evaluate the severity (9), occurrence frequency (3), and detectability (2), with the RPN (Risk Priority Number) being 54. Even if the RPN does not exceed the threshold, still change the ordinary screws to anti - loosening screws, and the RPN drops to 18.
3. Use the NPI (New Product Introduction) process to ensure the implementation of the design
NPI is divided into four stages, and each step must pass the "cross - departmental review":
T1 (Concept): Confirm that "this air conditioner features quiet operation, with a target noise level ≤ 30dB".
T2 (Detailed Design): Complete the drawings and BOM, and conduct prototype testing (measure the vibration value of the compressor).
T3 (Trial production): Produce 100 units in small batches. Conduct a 1000 - hour reliability test (simulating 3 - year usage). Mass production requires the approval of senior management.
T4 (Mass production): Solidify the design documents, train employees, and ensure that each unit meets the requirements.
This set of processes prevents functional problems from emerging. For example, a certain model of silent air conditioner has a complaint rate that is only one-third of the industry average after its launch - problems such as "compressor vibration", "fan noise", and "pipe resonance" have been solved during the design phase.
Supplier management: Quality is "bought"
"Quality is bought" - 30% of the defects come from suppliers' components. For example, an engine failure of an automobile enterprise was caused by the substandard material of the piston rings supplied by the supplier; the liquid leakage of the screen of a mobile phone enterprise was due to the insufficient compressive strength of the tempered glass.
The enterprise I serve has made supplier management a "systems engineering":
1. Access threshold: Focus on quality and capability rather than price
When selecting suppliers, the SQE (Supplier Quality Engineer) conducts a system audit: Check whether there is SPC for production monitoring, DFMEA for defect analysis, and SCAR (Supplier Corrective Action Report) for problem - solving. Only when the audit score is ≥80 (out of 100) can the supplier have the qualification to submit samples. For example, a hardware supplier failed the audit due to "no temperature monitoring for the injection molding machine" and was only allowed to enter after installing the SPC system.
2. NCI (New Component Introduction): Ensure the quality of samples
All new parts go through NCI (similar to automotive PPAP):
Dimension inspection: Measure each inch with a coordinate measuring machine, and it complies with the drawing.
Material testing: Use a spectrometer to measure the composition, and it does not contain harmful substances.
Performance test: Test the plastic's temperature resistance (no deformation after 24 hours at 100°C) and impact resistance (no cracking when dropped from a height of 1 meter).
3. SCAR system: Transparently monitor suppliers
The globally unified SCAR platform records all supplier issues. For example, if an electronic component supplier has had the problem of "pin oxidation" three consecutive times, and after investigation by the SQE, the cause is found to be "excessive humidity in the warehouse". The SQE helps the supplier install a dehumidifier (humidity ≤ 60%), and the problem is solved.
Manufacturing process: Use data and processes to "nip problems in the bud"
"Manufacturing quality is not about 'checking for defective products', but about 'preventing defective products from being produced,'" said a quality expert. The enterprise I serve controls manufacturing with "data + processes":
1. Find improvement directions from "poor quality cost"
Establish a COPQ (Cost of Poor Quality) system tois a Chinese word and the appropriate English word here should be "calculate" or "statistically analyze". Here we use "calculate", so it is"calculate the costs of BOM errors, warehousing damages, rework, returns, etc. For example, in the COPQ of a certain factory, "poor welding" accounts for 30%. Use DOE (Design of Experiments) to optimize the parameters - current 120A, time 3 seconds, pressure 8kg. The defective rate is reduced from 5% to 0.5%, saving 150,000 yuan per month.
2. SPC: Use data to "give early warnings"
Monitor key processes with SPC: For example, in the injection molding process, the mold temperature fluctuates within ±1℃. Once the threshold is exceeded, the machine will stop immediately. A production line stopped due to "mold temperature exceeding the upper limit". After investigation, it was found that "the cooling water pipe was blocked", thus avoiding 1000 defective products.
3. Foolproof design: It's hard to make a mistake
Use "Poka - Yoke" to prevent errors from occurring.
- Design the parts to have only one insertion direction to avoid incorrect installation.
- Install sensors on the equipment. If no parts are placed, the equipment will stop.
- Use "pictures + steps" in the work instruction to avoid misunderstandings.
For example, for the issue of missing screws on a certain assembly line, install a counting sensor – each time a screw is installed, the light lights up, and it only prompts completion after 4 screws are installed. As a result, the missing installation rate has dropped from 2% to 0.
Customer service: The "last mile" of quality is a closed loop
"The end - point of quality is service" - what customers buy is the "experience". For example, if a certain mobile phone has good quality but the customer service says "I don't know", the customer will think it has "poor quality"; if a certain air - conditioner has no problem but the customer has to wait for 3 days for the warranty service, the customer will turn into a negative critic.
The enterprise I serve has turned its services into a "closed-loop for quality improvement":
1. Warranty system: Identify the "quality pain points"
The globally unified warranty database records the reasons for the warranty of each product (such as "refrigerator compressor failure" and "air conditioner remote control malfunction"). It statistics the "Top 5 problems", for example, "the air conditioner filter gets dirty easily", and provides feedback to the design department - "Is the ease of cleaning not considered?" The complaints about a certain refrigerator with "lights not working" account for 15%. The design department replaces the light bulbs with LEDs (the lifespan increases from 1,000 hours to 50,000 hours), and the complaint rate drops to 0.1%.
2. 8D Report: Convert complaints into "improvement opportunities"
Major complaints are resolved using the 8D report:
- D1 (Form a Team): Design, quality, production, and supplier engineers form a team.
- D2 (Describe the problem): "The dimensions of this batch of parts are out of tolerance, and the customer cannot assemble them."
- D3 (Interim Measures): Recall the products and resend qualified parts.
- D4 (Finding the root cause): "Dimensional out-of-tolerance caused by mold wear";
- D5 (Permanent measure): Replace the mold and detect wear and tear quarterly.
- D6 (Verification): Produce 100 pieces. The measured dimensions are in line with the requirements.
- D7 (Prevention): Add "mold inspection" to daily maintenance.
- D8 (Closed): Provide feedback to the customer and confirm their satisfaction.
The power of the system: From "rule by man" to "rule by law"
The quality director of a Fortune 500 company said, "Excellent enterprises rely not on 'outstanding individuals' but on 'outstanding systems'." In that enterprise, even if the quality manager leaves, the system remains; even if the employees are replaced, the processes remain; even if the suppliers are changed, the audit standards remain.
In contrast, for many Chinese enterprises, product quality depends on "personal capabilities". When there is a capable quality manager, the product quality is good; when he leaves, the quality declines. Experienced veteran workers can spot defective products; when they retire, the defective product rate goes up. This is the difference between "rule by man" and "rule of law". A system enables ordinary people to do extraordinary things, while rule by man only allows extraordinary people to do ordinary things.
Conclusion: The "Right Way" to Upgrade Quality
The quality problem of Chinese enterprises is not due to "lack of attention" but "failure to find the right method" — not by "adding inspectors" or "punishing workers", but by "building systems"; not by "making hasty decisions" or "relying on experience", but by "using data" and "following procedures".
The success of that Fortune 500 company is a "systematic success" – from design to suppliers, from manufacturing to services, every link has "implementable processes", "measurable indicators", and "traceable responsibilities". This system is not "fabricated"; it is "summarized from experiences of making mistakes". It is not "unchanging" but "improved annually".
For Chinese enterprises, there is no shortcut in quality upgrading, only foolish methods: refine the quality of each link, tighten the standards of each process, and find out the root cause of each problem. Only when the system replaces the individual as the quality guarantee can enterprises truly get out of the fire - fighting dilemma and achieve a fundamental improvement in quality.