During the tea break, a colleague sitting next to me suddenly put down his coffee cup, stared at me and asked, "The new company has just started, and even the processes haven't been sorted out. How can we build a quality management system?" His question hit a blind spot for many entrepreneurs - they always think that quality management has to be "high - end and sophisticated", and that they have to first develop system documents and conduct full - staff training, but forget that for a new enterprise, the most urgent thing is to "survive": first hold the bottom line of quality, and then talk about optimization.
My first suggestion to him at that time was to "return to the origin of quality management": quality is "inspected out". This statement may be outdated for established enterprises, but for new enterprises, it is the most practical starting point. New enterprises lack resources, processes, and experience and cannot implement "preventive management" like large enterprises (for example, optimizing the manufacturing process with Six Sigma). The top priority is to first block the "visible mistakes". The so - called "origin" means to first control "both ends": one end is the incoming raw materials, and the other end is the outgoing finished products.
First, let's talk about "quality control at the feeding end". The raw material quality of new enterprises often fluctuates the most. The suppliers may also be small factories, or the cooperation has just started, and no stable quality agreement has been established. At this time, "precise inspection" should be carried out instead of "full inspection".Step 1: First, define the "bottom line that must be adhered to" (referred to as Critical-to-Quality Characteristics, CTQ in the industry). For example, when manufacturing electronic components, the pin spacing of chips and the withstand voltage value of capacitors should be determined. When making food, the moisture content of flour and the limit of additives should be determined.Step 2: Select the sampling rule. Use the AQL (Acceptable Quality Level) 1.0 standard, sample 125 samples. If the number of non-conforming products does not exceed 3, the products can be released, which saves manpower and controls risks.Step 3: Implement the handling of non-conforming products. If the IQC (Incoming Quality Control inspector) finds that the melt index of plastic pellets fails to meet the standard, directly initiate the return process and never let defective raw materials enter the production line. At the beginning, the process can be simplified. Ask the supplier to provide the COA (Certificate of Analysis), and then randomly test 1 - 2 key items, which can quickly inspect the factory and avoid the waste of full inspection.
Now let's talk about "quality control at the shipping end". The customers of new enterprises are like "probationary employees". One instance of substandard products can shatter trust. Shipping inspections should be aligned with the real needs of customers. For example, if a customer requires "scratch-free phone cases", the definition of "scratch" must be clearly defined: length ≤ 2mm, depth ≤ 0.1mm, and not in the front visible area. Then conduct the final quality control (FQC): make a checklist, and tick and confirm each item from appearance, function, packaging to labeling. More importantly, it is batch traceability - assign a batch number to each finished product and associate it with the raw material batch, production workshop, and operator. In case the customer reports substandard products, the problem batch and the step where the problem occurred can be traced within 2 hours, enabling quick product recalls or replenishments. I once helped a new toy factory with shipping inspections and required each batch of toys to be affixed with a "batch QR code". Later, when the heavy metal content in the paint of a certain batch exceeded the standard, 300 pieces were directly recalled without affecting other customers, thus preserving the orders.
Finally, there is "iteration from inspection to the system". Once the processes at both "ends" are stable, use data to drive improvements. For example, if the IQC records show that a certain plastic supplier often has dimensional deviations, don't just return the goods - go to the workshop with the supplier to see if the temperature of the injection molding machine is not well - controlled or if the mold is worn out, and help him optimize the manufacturing process. If the FQC records show that the finished products on a certain production line often have "non - functioning buttons", then observe the operator's techniques to see if the soldering temperature is insufficient, and then optimize the SOP (Standard Operating Procedure) by adding a "temperature calibration" step. Gradually, the inspection data will become "prevention clues": from "blocking defective products" to "reducing defective products", and then to "avoiding defective products". I once worked with a new electronics factory. At first, it only carried out IQC and FQC. Later, it used the inspection data to identify 3 high - frequency problems, optimized the manufacturing processes of 2 suppliers, and adjusted the parameters of 1 production line. After half a year, the defective rate dropped from 5% to 1%. At this time, introducing the ISO9001 system is not just for show - every clause in the system is summarized from actual problems.
The logic of building a quality system for a new enterprise is like "building a house": first, lay the foundation (incoming and outgoing inspections), then build the walls (solve high-frequency problems), and finally put on the roof (improve the system). Don't start with abstract "quality policies" and "quality manuals". First, safeguard the "two ends", and then iterate with data. Only in this way can quality management be transformed from a "slogan" into a "money-making tool". After all, for a new enterprise, "good quality" is not the goal; "surviving and making more money" is.