QM Management Depth (9) | Four-Class Calculation Method for Quality Costs: Prevention, Appraisal, Internal Failure, External Failure

A certain automotive parts company, with annual revenue of 1.2 billion yuan, supplies two original equipment manufacturers (OEMs). In April 2025, the quality director presented the quality cost report to the management meeting for the first time. The report was prepared by the finance department according to expense categories, and the numbers were very "clean": prevention costs accounted for 0.3% of revenue, appraisal costs 3.8%, internal failure costs 1.0%, and external failure costs 0.1%, totaling 5.2%, or about 62 million yuan.

The general manager quickly drew a conclusion: "Appraisal costs account for more than 70%, which indicates that we rely on inspections to ensure quality. This efficiency is too low." The meeting decided on the spot to reduce the final inspection personnel by 15% within three months, with 12 people being reassigned to production. The quality director did not stop this decision because he had no second piece of evidence—while he knew something was wrong with the numbers, he could not pinpoint the issue.

Three months later, in July, customer complaints increased from an average of 3 per month to 9, with one incident causing a 4-hour shutdown at an OEM. This time, he brought evidence to the meeting: the 3.8% appraisal cost included the working hours of 14 quality engineers for quality planning, training, supplier coaching, and poka-yoke implementation, amounting to about 4.6 million yuan. The actual appraisal cost was only 2.5%, and the prevention cost was actually 1.6% instead of 0.3%. The company was not "inspection-heavy and prevention-light," but the accounting was incorrect; the company cut not redundancy but a line of defense.

1. The Essence of the Problem: Calculation Method is a Governance Design, Not a Statistical Technique

Most quality managers treat quality cost calculation as a technical task: getting numbers from finance, applying formulas, and generating reports. The real challenge has never been in the formulas but in the following three aspects.

Perspective One: The Definition of Scope is the Setting of the Agenda

The boundaries of the four cost categories are artificially drawn. For the same "quality engineer salary," whether it is counted as prevention or appraisal has no accounting standard to follow, only business judgment. Whoever defines the scope decides what story the report tells. If the finance department defines the scope, the story is "which department spent how much money," naturally leading to cost-cutting. If the quality department defines the scope, the story is "whether the money was spent before or after defects," which may lead to structural adjustments. When the boss only looks at one page of a PPT, this difference can determine the budget's fate.

Perspective Two: Errors in Scope Definition Can Be Amplified into Incorrect Decisions

Every structural misjudgment in the quality cost report is translated into an action at the management meeting. Underestimating prevention costs leads to the conclusion that "prevention does not need investment"; overestimating appraisal costs leads to the conclusion that "inspections can be reduced"; and failing to record internal failures leads to the conclusion that "quality is still good." To some extent, the calculation is producing facts—the cost structure seen by management is the cost structure they base their decisions on. This is the fundamental reason why the calculation system must be led by the quality department, not because the quality department wants to take over tasks, but because only the quality department has a comprehensive understanding of both "the process in which activities occur" and "the results of the expenses incurred."

Perspective Three: Granularity is Not Always Finer the Better, but Depends on the Target

Calculating by product line targets pricing and order selection; by process targets process improvement; by customer targets the allocation of quality investment; and by responsible department targets performance evaluation. Four levels of granularity correspond to four types of decisions. Trying to do all at once results in four inaccurate tables, and ultimately, none of them will be used. A rational approach is to start with two dimensions in the first year: a total table and a product line breakdown table, attributing 80% of the costs to 80% of the problem sources, and then gradually adding more dimensions.

2. Practical Actions: Five Operational Steps

Step One: Divide Calculation Responsibilities into "Three Segments." Data is reported by the departments where it occurs (scrap reports, rework hours, claim forms, overtime and freight documents). The quality department verifies and categorizes the data (determines the归属 based on the boundaries of the four categories and generates the calculation table). The finance department confirms the accounts (maps the categorized results to financial categories and verifies the amounts and periods). Criterion: Each monthly calculation table must have signatures from all three segments—tables signed by the quality department alone are not credible in the long term.

Step Two: The Scope Must Be Documented and Controlled for Release. Create a "Quality Cost Categories and Scope Definition Table": the left column lists the four cost categories, and the right column lists specific categories, definitions (what counts, what does not), rate criteria, data sources, and responsible persons. The scope table must be signed by the quality director and the finance manager, included in controlled documents, and any changes must be re-approved. Criterion: Select any three categories, and the independent categorization results of two calculation personnel should be consistent.

Step Three: Use Work Coefficients to Solve the "Multi-Role" Time Allocation. The time of quality engineers (QEs), inspection team leaders, and quality managers spans both prevention and appraisal. Counting the entire time as appraisal is a direct cause of underestimating prevention costs. The solution is to establish a work coefficient table for cross-category positions, initially based on the actual distribution of working hours over a week (for example, quality engineers are split 0.6 for prevention and 0.4 for appraisal), and to keep monthly working hour records as evidence, with coefficients calibrated every six months. Criterion: After allocation, the proportion of prevention costs in total quality costs should fall within the 10% to 15% empirical range—long-term proportions below 5% generally indicate a problem with the scope, not that the company really does not need prevention.

Step Four: Fix the Calculation Table Template, Each Table Targets One Decision. The total table shows the proportions and trends of the four categories (for the management meeting). The product line breakdown table shows where losses are concentrated (for orders and resource allocation). The process or workshop allocation table shows the priority of process improvements (for process and production). The external failure detail table shows customer and responsibility boundaries (for customers and suppliers). Criterion: Each breakdown table should point to a clear action—columns that do not point to actions should be deleted to avoid occupying reporting costs.

Step Five: Calculation Results Must Be Linked to Management Meeting Resolutions. Generating the table is just the beginning. Set thresholds (for example, a category exceeding the budget by 10% or the four-category structure deviating from the target line), and when triggered, form a written resolution: the responsible person, amount, and completion deadline must be clearly stated. Criterion: Within five working days after the table is generated, each category exceeding the limit should have a corresponding resolution record. If this cannot be achieved, the calculation system will degrade into a monthly statistic, and no one will change their behavior because of it.

3. Case Development: Scope Adjusted for Two Quarters, Decision Menu Changed

Returning to that company. After the shutdown incident in July, the first correct action the quality director took was not to rush for data but to reclaim the scope definition. He spent two afternoons with the finance manager, breaking down the working hours of 14 quality engineers by work coefficients. Planning, training, supplier coaching, and poka-yoke implementation were counted as prevention, while daily inspections and tests were counted as appraisal. Rework hours were priced based on the workshop hourly cost, and shutdown losses were calculated based on the marginal profit per minute. Customer claims, on-site service travel, and in-house handling of customer complaints were all counted as external failure, no longer remaining in internal failure.

The process was not smooth. The production manager directly questioned in the meeting: pricing rework hours based on the workshop hourly cost, which includes equipment depreciation and allocation, would "inflate" the loss. This argument actually holds, as the marginal resource input for rework only includes direct labor and energy consumption. The management meeting ultimately ruled to use a marginal perspective, reducing the rework hourly rate by about 35%, which both parties accepted. This concession was valuable—once the scope is questioned and no one responds, the report is immediately categorized as "numbers made up by the quality department."

The second action was to add personnel. The quality department added 1 position for cost calculation and analysis, with the finance department investing about 0.4 full-time equivalents (FTEs) per month and the workshop filling out three additional forms each month. The additional personnel cost for one year was about 2.5 million yuan. The third action was to fix the table release schedule to the 8th of each month, with the table being reviewed in the management meeting within one week of release, and a resolution formed for each category exceeding the limit.

By the following June, the structure had changed: prevention costs increased from 0.3% to 1.5%, appraisal costs decreased from 3.8% to 2.3%, internal failure costs decreased from 1.0% to 0.4%, and external failure costs decreased from 0.1% to 0.08%, totaling a reduction from 5.2% to 4.3%, or about 10 million yuan. Among the 15 reassigned personnel, 9 returned to quality positions: 5 to supplier quality and process quality, and 4 to final inspection positions.

The cost is also evident. The 1.8 million yuan claim from the July shutdown could not be recovered, and the customer demoted the company to a backup in the next year's new project bidding, resulting in a loss of about 40 million yuan in orders—this is the real price of "incorrect reports leading to incorrect decisions," more expensive than any training course. The quality director's original words in the year-end management review were: "If the scope is wrong, the more frequently you calculate, the further off you get."

The last two rules. First, once the scope is released, it should only be calibrated once a year, and not adjusted temporarily because the numbers for a certain quarter look bad—once the trend line breaks, the calculation system loses all its persuasiveness. Second, do not directly link the calculation results to the department's current bonus. Once linked, the reporting phase immediately becomes a game: scrap reports are delayed, and rework hours are underreported by half an hour, making the report look good but completely useless. Its purpose is to allocate resources, not to distribute bonuses.

4. Self-Inspection Checklist

• The calculation table is signed by "reporting from the departments where it occurs, categorization by the quality department, and confirmation by the finance department," not just numbers from the quality department.
• There is a controlled and released "Quality Cost Categories and Scope Definition Table," and the categorization results of two calculation personnel for any three categories are consistent.
• The working hours of cross-category positions (QEs, inspection team leaders, quality managers) are allocated by work coefficients, and the proportion of prevention costs is close to the 10% to 15% empirical range.
• Each breakdown table can point to a clear action (pricing, improvement, customer investment, or supplier responsibility).
• A written resolution, including the responsible person, amount, and completion deadline, is formed for each category exceeding the limit within five working days after the table is generated.

If any two of the above five points are not met, it indicates that the current calculation is still at the "financial statistical report" stage. First, supplement responsibilities and scope, then discuss trends and structure—a report that no one acknowledges and that does not lead to any action is just a cost, no matter how accurate it is.

If the scope is wrong, the more frequently you calculate, the further off you get.

Knowledge code: 4.3.1

Version: v20260919

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.