How Many People Should the Quality Department Have? —— A Five-Step Method for Staffing Calculation from "Output Ratio" to "Workload Hours"

At the beginning of the year budget meeting, the finance department insisted that the quality personnel should not exceed 1% of the total workforce. However, the report from a second-party customer audit stated, "Inspection manpower is insufficient, posing a risk of missed inspections." The quality manager, caught in the middle, could only say, "We are really busy." But how busy? They couldn't quantify it. When the boss asked, "How many people should we have?" the response was, "The industry standard is 3% to 5%." The boss then asked, "Why should we go with 5%?" and the manager was at a loss for words.

This is a staffing issue, not a personnel addition issue. The essence of staffing can be summed up in one sentence: Convert the workload of a department into verifiable hours, then divide by the effective hours one person can provide. After the calculation, you will realize two things: first, the ratio method can give absurd answers in both directions; second, what often needs to be done first is not adding more people but reducing the workload that should not be borne by humans.

This article provides a directly applicable calculation process: a list of six types of quality activities, an inspection workload calculation form, a QE workload table, a staffing structure table (core pool + flexible pool), and a cost criterion for determining whether to reduce personnel.

1. Why the Ratio Method Always Gets It Wrong

There are two versions of the quality personnel ratio commonly used in the industry: by the total workforce (3% to 8%) and by output value (2 to 3.5 people per 100 million yuan of output value). These numbers are useful for post-event verification but will certainly be incorrect if used as a basis for staffing. The reason is that they treat "output value" as a proxy for "inspection demand," but there is no stable relationship between the two.

Three typical failure scenarios:

Changes in Outsourcing Ratio. A company reduced its in-house production rate from 70% to 45%. The output value did not increase, but the number of incoming batches rose from 320 per month to 620. The workload for incoming inspection is determined by the number of batches and the sampling quantity, not by the output value. According to the output value ratio, the staffing does not need to change; however, based on the actual workload, IQC needs 6 more people.

Changes in Product Structure. The inspection time for a single precision machined part is approximately 4 to 6 minutes (multiple dimensions, multiple tolerances, requiring specialized inspection tools), while a total assembly part might only take 40 seconds. For the same output value of 100 million yuan, the former requires more than five times the inspection hours of the latter.

Differences in Customer Requirements. A customer specifies in the quality agreement, "100% inspection of key characteristics, and an inspection report for each batch." This immediately triples the inspection demand for the production line. Such requirements do not reflect in the output value but manifest as overtime for a specific team.

Therefore, the ratio method can only be used for one thing: compare the calculated result with industry peers. If the deviation exceeds one-fold, investigate the cause. It cannot be used to decide staffing because it does not differentiate "what to inspect, how much to inspect, and how long to inspect."

2. The Underlying Formula and Three Staffing Pools

Demand Hours ÷ Effective Hours per Person = Required Number of People.

Many people miscalculate the effective hours per person. It is not 21.75 days × 8 hours = 174 hours, but rather, it should be multiplied by an activity rate: inspectors have non-inspection tasks such as material requisition, sample submission, waiting, record organization, pre-shift meetings, and equipment checks, plus absences due to leave. An industry-stable activity rate of 0.85 is commonly used, meaning each person has about 148 effective inspection hours per month. If your record-keeping volume is particularly high (paper three-inspection forms, manual ledgers), this rate should be reduced to 0.75 to 0.80, not forced to 0.85.

Staffing should be divided into three pools, otherwise, it will be unclear:

A constraint often overlooked in the fixed pool is that it cannot be linearly converted based on workload. For a company with two sites and three system certifications (ISO 9001 + IATF 16949 + customer-specific requirements), having fewer than 1.5 people in the system position will lead to issues such as expired documents not being reviewed, reduced internal audit plans, and incomplete management review inputs. The cost of these issues will not be immediately apparent but will be settled at the next certification audit.

3. Five-Step Calculation: From Activity List to Staffing Structure

Step One: Inventory Six Types of Quality Activities, Not Starting from Positions

The starting point for calculation is not "what are the current people doing," but "what quality activities does this company need." Inventorying from positions will treat historical tasks as necessary tasks; inventorying from activities will help identify self-created tasks.

After inventorying, add two numbers for each activity: single-instance hours and monthly frequency. These numbers should not be guessed; instead, have three people record their actual time for a week and take the median. The most underestimated inspection hours are often "non-inspection actions" such as finding drawings, waiting for cranes, re-measuring, issuing nonconforming product reports, and discussing judgment standards with production, which can account for 30% to 40% of the total time per piece.

Step Two: Calculate Inspection Hours

Taking an example of an automotive parts company (annual output value of 820 million yuan, monthly production of 120,000 finished products, 6 production lines in two shifts, 85 incoming suppliers, 14 ongoing new projects). The current quality department has 14 people: 9 inspectors, 4 QEs, and 1 system engineer.

2,092 ÷ 148 ≈ 14.1 people.

This result itself is the first discovery: the current 9 inspectors can only provide about 1,331 hours based on effective hours, leaving a shortfall of 761 hours. Spread over 9 people, this is an average of 84.5 hours of overtime per person per month, equivalent to working an additional 10.5 shifts per month. This explains why the phrase "we are very busy" is vague—it can actually be quantified to 761 hours.

Step Three: Calculate QE and Fixed Load

QE hours cannot be calculated based on production volume but must be based on events. The QE load for the same company:

959 ÷ 148 ≈ 6.5 people. Currently, there are 4 people, leaving a shortfall of 2.5 people. The fixed pool (system + metrology + training and certification) should have a minimum of 1.5 people, but currently, there is 1 person.

Step Four: Summarize and Put the Shortfall on the Table

At this point, the nature of the staffing negotiation changes. Originally, the debate was about "whether to add more people," but now it is about "which part of the 2,092 hours should not exist." The latter can be discussed, while the former can only be decided by the budget.

Step Five: Reduce Workload First, Then Adjust Staffing

Directly taking the calculation result to request 8 more people will likely be rejected. The correct sequence is to first reduce the workload, compress the "work that must be done by people" to the minimum, and then request the remaining people.

Total reduction: 564 hours. Inspection workload drops to 1,528 hours, requiring 10.3 people. The new staffing structure becomes: inspection 10.3 + QE 6.5 + fixed 1.5 = 18.3 people, with a shortfall of 4.3 people compared to the current 14 people. This number can be implemented over two years: add 2 QEs in the first year (complaints and supplier coaching are the most urgent current shortfalls), and add 2 inspectors in the second year.

Peak periods should not be addressed by staffing but by the flexible pool. The flexible pool consists of three parts: cross-trained personnel (production team leaders, process engineers trained to obtain inspection qualifications, totaling 6 people), third-party sorting and on-site inspection, and external laboratories. The flexible pool should have trigger lines written into the department management documents to avoid long-term use:

• Average overtime exceeds 12 hours/month for 2 consecutive months;
• Monthly complaints exceed 4 cases;
• Concurrent new projects exceed 16;
• Temporary stringent requirements from a single customer (full inspection, severe sampling).

Any trigger will activate the flexible pool, and the exit time should be explained at the monthly quality meeting.

4. Five Common Pitfalls to Avoid

Pitfall One: Using Average Output Value to Prove "High Efficiency." High efficiency can have two meanings: genuinely high efficiency or missed inspections. The basis for judgment is not the average output value but whether external failure costs (complaint handling, claims, returns, sorting) have increased concurrently. If efficiency has improved year after year but external failure costs are also rising, it is not efficiency but missed inspections.

Pitfall Two: Using "Low Inspection Defect Rate" to Prove Too Many Inspectors. The denominator for the inspection defect rate is the quantity inspected, not the quantity missed. To prove inspection effectiveness, a blind sample challenge should be conducted: insert known defective items into normal batches and see if the inspectors can identify them. If the identification rate is below 90%, it indicates that the current inspection pace has exceeded human capability, and reducing personnel is a pure gamble.

Pitfall Three: Assigning Efficiency Metrics to Inspectors. Once "average pieces inspected per person" becomes an inspector's performance metric, the most rational action is to inspect less and speed up the pace. The performance metrics for inspection positions should be the missed inspection rate (blind sample challenge) and judgment consistency (attribute consistency analysis), not the number of pieces.

Pitfall Four: Focusing Only on Staffing Savings, Not on Increased Failure Costs. The annual comprehensive cost for one inspector is about 120,000 to 150,000 yuan; the handling cost for one complaint (transportation, sorting, rework, customer fines, on-site personnel) is typically 8,000 to 20,000 yuan, with major claims calculated separately. The criterion is: whether the estimated increase in external failure costs after reduction is less than the staffing savings. A company reduced IQC from 6 to 4 people, saving 260,000 yuan annually, but the next year, customer returns and sorting costs due to incoming material issues amounted to 710,000 yuan—saving the budget but losing profit.

Pitfall Five: Treating Third-Party Sorting as a Long-Term Solution. Third-party sorting is charged hourly, cannot be budgeted, and has unclear responsibility boundaries. More importantly, it does not build any internal capability: the same issue will be sorted again next year. It is only suitable as a containment measure, not as a substitute for proper staffing.

Another Easily Overlooked Point: Ignoring Project Peaks. During the introduction period of new projects, the monthly workload for QEs is 2 to 3 times the steady state. Staffing based on steady-state load will lead to all personnel being overwhelmed when new projects come in, causing steady-state work (complaints, supplier coaching) to be interrupted. The cost of such interruptions usually returns in the form of batch issues after two to three months. Staffing should be designed in two segments: "steady-state load + flexible pool during peak periods," rather than using the annual average.

5. Rules for Resolving Cross-Functional Conflicts

Staffing issues are often unresolved because each party speaks a different language: production wants more people (capacity priority), finance controls headcount (cost priority), and customers want coverage (risk priority). Negotiating based on headcount will never reach a conclusion.

A useful rule for resolution is: Inspection density is determined by the number of special characteristics, not by headcount. The specific approach is to count the number of key characteristics (SC/CC) for each product from the control plan, determine the inspection method (full inspection, sampling inspection, exemption) and sampling level based on the number of characteristics and the consequences of failure, and then back-calculate the hours. This way, "whether to add an inspector to this line" becomes a technical issue: this line has 14 key characteristics, the current sampling level is insufficient, and the blind sample challenge identification rate is only 78%—the conclusion is naturally established, without the need for persuasion.

6. In One Sentence

Staffing is not about headcount but about converting "busyness" into hours—first, calculate the 2,092 hours, then reduce the 564 hours that should not be borne by people, and finally, add the truly missing few people.

Staffing depends on hours, not on ratios.

Knowledge code: 13.1.2

Version: v20261026

Author: QTank QTank is dedicated to providing systematic professional knowledge, methodologies, and practical tools for quality management practitioners, helping companies continuously improve their quality capabilities.