Procurement Glossary
Sampling: definition, methods and application in Procurement
November 19, 2025
Sampling is a central quality assurance procedure in Procurement in which only a representative part of a total quantity is tested. This method enables companies to monitor quality standards cost-effectively and evaluate supplier performance. Find out below what sampling means, which methods are used and how you can use them strategically in the procurement process.
Key Facts
- Statistical test method for quality assessment without full testing
- Reduces testing costs by up to 80% with the same informative value
- Based on mathematical models such as AQL (Acceptable Quality Level)
- Enables continuous supplier evaluation and risk minimization
- An integral part of modern quality management systems
Contents
Definition: sampling
Sampling is a statistical procedure for quality control in which a representative subset of a population is selected and tested.
Basic principles of sampling
The procedure is based on statistical principles and allows conclusions to be drawn about the overall quality of a delivery. The central elements are
- Sample size according to statistical specifications
- Random or systematic selection of test objects
- Defined acceptance and rejection criteria
- Documentation and evaluation of the test results
Random sample inspection vs. full inspection
In contrast to the full inspection, not all units are inspected, but only a statistically relevant proportion. This significantly reduces costs and inspection time, while the quality inspection still delivers meaningful results.
Importance in Procurement
For purchasing organizations, random sample testing is indispensable for efficient delivery quality controland for evaluating supplier performance as part of quality management in Procurement.
Methods and procedure for sampling inspections
The test is carried out according to standardized procedures, which can vary depending on the test objective and type of material.
AQL sampling method
The Acceptable Quality Level (AQL) defines the maximum acceptable error rate in a delivery. The incoming goods inspection based on AQL sampling is carried out according to defined tables that specify the sample size and acceptance figures.
Statistical process control (SPC)
SPC uses control charts to continuously monitor processes. Samples are taken at regular intervals and statistically evaluated in order to identify trends and deviations at an early stage.
Test equipment management and MSA
Professional test equipment management is required for reliable results. The measurement system analysis(MSA) ensures that the test equipment used has the required accuracy and repeatability.
Tacto Intelligence
Combines deep procurement knowledge with the most powerful AI agents for strong Procurement.
Important KPIs and targets
Key figures for sampling inspections make it possible to evaluate the inspection efficiency and quality performance of suppliers.
Inspection cost ratio and efficiency
The relationship between inspection costs and the value of the goods shows the cost-effectiveness of the random sample inspection. Typical target values are 0.5-2% of the value of the goods. In addition, the inspection time per batch is measured in order to optimize throughput times and reduce quality costs.
Error detection rate
This KPI measures how many defective lots are actually identified by the random sample inspection. Values above 95% are considered very good. The rework rate shows the proportion of lots that have to be reworked due to sampling results.
Supplier performance indicators
PPM values (parts per million) from random sample inspections objectively assess supplier quality. Acceptance rates and blocked stock ratios are also recorded in order to optimize blocked stock management.
Risks, dependencies and countermeasures
Despite its advantages, sampling involves specific risks that must be minimized by taking appropriate measures.
Statistical uncertainty
Samples can never offer 100% certainty. Defective batches can remain undetected if defective parts are not included in the sample. Appropriate sample sizes and supplementary control plans at the supplier's premises provide a remedy.
Test equipment errors and measurement uncertainty
Inaccurate or poorly calibrated test equipment distorts results and leads to incorrect decisions. Regular calibration and Gage R&R studies ensure the reliability of the measurements.
Supplier dependency
Inadequate quality systems at the supplier increase the risk of quality problems. Supplier audit programs and layered process audits create transparency and promote continuous improvement.
.avif)
Practical example
An automotive supplier receives 5,000 screws from a supplier every day. Instead of a full inspection, a random sample of 200 screws is taken in accordance with AQL 1.5 and checked for dimensional accuracy. If a maximum of 7 defective parts are found, the entire delivery is accepted. If more than 7 defects are found, a 100% sorting or return is carried out. This procedure reduces the testing effort by 96% while maintaining a high level of statistical reliability.
- Sample size: 200 out of 5,000 parts (4%)
- Acceptance number: maximum 7 faulty parts
- Time saving: 95% less testing time
- Cost reduction: 80% lower testing costs
Trends & developments in sampling inspections
Modern technologies and digital transformation are fundamentally changing sampling and opening up new possibilities.
AI-supported sample selection
Artificial intelligence optimizes the selection of samples by analyzing historical data and supplier performance. Machine learning algorithms identify patterns and dynamically adjust sample sizes to risk profiles, increasing the efficiency of the quality assurance agreement.
Digital test documentation
Cloud-based systems enable the real-time recording and evaluation of inspection data. Integration into ERP systems creates end-to-end traceability and automated reporting for supplier evaluations.
Predictive quality analytics
Predictive analyses use sample data to forecast quality problems. This development supports proactive measures within the framework of quality gates and reduces failure risks in the supply chain.
Conclusion
Sampling is an indispensable tool for efficient quality assurance in Procurement. It enables cost-optimized quality control with statistically verified validity and supports continuous supplier evaluation. Modern digital technologies and AI approaches significantly expand the possibilities and create new potential for forward-looking quality assurance. The strategic use of spot checks makes a significant contribution to minimizing risk and optimizing costs in procurement.
FAQ
What is the difference between AQL and other sampling methods?
AQL (Acceptable Quality Level) defines the maximum acceptable error rate and uses standardized tables for sample size and acceptance numbers. Other methods such as variable sampling or continuous sampling have different statistical bases and areas of application depending on the inspection objective.
How do I determine the optimal sample size?
The sample size depends on the batch size, the desired AQL level and the inspection level. Standard tables in accordance with ISO 2859 or MIL-STD-105 provide specific guidelines. Larger samples are selected for critical characteristics, while smaller samples may be sufficient for non-critical characteristics.
What are the costs of random sampling?
The main costs include personnel costs for testing, test equipment costs, documentation and any rework. Typically, the total costs are 0.5-2% of the value of the goods. This is offset by the costs saved by avoiding full inspections and reducing quality problems.
How do I deal with rejected lots?
If a lot is rejected, several options are available: Return to supplier, 100% sorting to reject defective parts, rework at supplier or use with deviation approval. The decision depends on criticality, costs and delivery dates.
.png)
.png)
.png)
