QC, or quality control, is the inspection side of quality management. It refers to the operational techniques and activities used to verify that a product or service meets defined requirements. In testing specifically, QC is the hands-on work of checking outputs, catching defects, and confirming that what’s been built actually works as intended.
Whether you’re in software development, manufacturing, or laboratory science, QC testing follows the same core principle: examine the end result, compare it against a standard, and flag anything that falls short.
How QC Differs From QA
QC and QA (quality assurance) get used interchangeably, but they cover different ground. QA is about process: setting up the right workflows, documentation, and standards so that quality is built in from the start. QC is about verification: inspecting the actual product to confirm those standards were met. Think of QA as designing a recipe and QC as tasting the dish.
QA covers virtually all of a quality system, while QC is a subset of those activities. Auditing, for example, falls under QA. Running a test suite against finished code, pulling samples off a production line, or checking lab results against control values are all QC. QA is proactive and preventive. QC is reactive and detective. Both are necessary, and in practice they overlap constantly, but understanding the distinction helps you know which problems to solve with better processes versus better inspections.
QC in Software Testing
In software, QC is the actual execution of tests to find bugs before users do. It spans multiple levels, each catching a different class of problem.
- Unit tests are the most granular. They check individual functions or methods in isolation, verifying that a single piece of logic behaves correctly with known inputs.
- Integration tests verify that separate modules or services work together properly. Testing whether your application communicates correctly with its database is a classic integration test.
- System tests evaluate the complete application as a whole, simulating real user scenarios end to end.
- Regression tests re-run previous test cases after code changes to make sure new updates haven’t broken existing functionality.
Most software teams automate a large portion of this work. Tools like Selenium handle web application testing, Appium covers mobile apps, and Postman specializes in API testing. Jenkins ties these tools into continuous integration pipelines so that QC checks run automatically every time a developer pushes new code. The goal is to catch defects as early as possible, when they’re cheapest and fastest to fix.
Manual QC still plays a role, especially for exploratory testing where a human tester deliberately tries unexpected inputs and workflows that automated scripts wouldn’t cover. The combination of automated and manual QC gives the most complete picture of product quality.
QC in Manufacturing
On a production line, QC means inspecting raw materials, in-process components, and finished goods against specifications. This can range from simple visual inspection to sophisticated statistical methods.
Statistical process control (SPC) is one of the most widely used QC techniques in manufacturing. It applies statistical analysis to monitor a production process in real time, flagging when measurements drift outside acceptable limits before they produce defective units. SPC uses control charts, sampling plans, and tools like defect maps and process flowcharts to keep production stable.
Teams track several key metrics to gauge how well QC is working. Defect density measures the number of defects per unit of production, calculated by dividing total defects by the number of units produced. First pass yield measures the percentage of units that pass all quality checks without any rework, giving a direct read on process efficiency. A high first pass yield means fewer resources wasted on corrections.
QC in Laboratory Testing
Clinical and research laboratories have their own rigorous QC protocols. Labs run control samples, materials with known values, alongside patient or research samples every day. If the control results fall outside expected ranges, something is wrong with the equipment, reagents, or process, and results cannot be released.
Labs use control charts called Levey-Jennings charts to plot daily control values over time. A set of decision rules known as Westgard rules determines when a result is acceptable and when it signals a problem. High-performing tests (those with very tight precision, rated at six sigma or above) may only need one control level run once per day. Less precise tests need two control levels run twice daily with multiple rules applied before results can be reported. Any test performing below three sigma, the minimum acceptable threshold, requires a root cause investigation before it can be used for patient results at all.
Standards and Regulations
QC isn’t just good practice. In many industries it’s a legal or contractual requirement. ISO 9001:2015 is the international standard for quality management systems, and it requires organizations to plan, implement, and control the processes necessary to meet customer requirements. Certification to ISO 9001 signals to customers and partners that a company has documented QC processes in place.
In regulated industries the requirements are even more specific. Medical device manufacturers in the United States must comply with FDA regulations under 21 CFR Part 820, which mandates a documented quality management system, detailed record-keeping for complaints and servicing, and traceability through unique device identification. Every test, inspection, and corrective action must be recorded. These records are subject to FDA audit at any time.
What Makes QC Effective
Regardless of industry, effective QC shares a few characteristics. First, it’s based on clear, measurable criteria. “Good enough” isn’t a specification. Teams need defined thresholds: acceptable tolerance ranges, pass/fail criteria, expected performance benchmarks. Without these, QC becomes subjective and inconsistent.
Second, effective QC happens at multiple stages, not just at the end. Catching a defect during a unit test costs minutes. Catching it after release costs hours or days of rework, plus the reputational damage. The same principle applies in manufacturing: inspecting incoming materials prevents defective components from entering the production line in the first place.
Third, QC generates data that feeds back into improvement. Every defect found is information about where a process is weak. Tracking defect density over time, monitoring first pass yield trends, and reviewing control chart patterns all point toward root causes that QA processes can then address. QC without this feedback loop is just sorting good from bad. QC with it becomes a driver of continuous improvement.