How To Calculate Average Outgoing Quality

Use this premium AOQ calculator to estimate the average outgoing quality of a lot under acceptance sampling with rectification. Enter your lot size, sample size, incoming defect level, acceptance number, and model to instantly calculate AOQ, probability of acceptance, outgoing defectives, and a visual AOQ curve.

Quality Engineering

Acceptance Sampling

AOQ and AOQL Analysis

AOQ Calculator

Average outgoing quality is commonly estimated with rectifying inspection as: AOQ = P_a × p × (N – n) / N, where rejected lots are screened and defectives are removed.

Core Formula P_a × p × (N – n) / N

Typical Use Rectifying inspection plans

Tip: For single sampling plans, binomial and Poisson methods are common approximations. If your organization already knows the probability of acceptance from an OC curve or software package, select the direct mode and enter Pₐ directly.

Expert Guide: How to Calculate Average Outgoing Quality

Average outgoing quality, often abbreviated as AOQ, is one of the most important measures in acceptance sampling and production quality control. It tells you the expected proportion defective in product after an inspection plan is applied, assuming rejected lots are screened and defective items are replaced or removed. In practical terms, AOQ helps manufacturers, suppliers, and procurement teams estimate what customers are likely to receive after sampling and rectification.

If you work in manufacturing, medical devices, electronics, food processing, packaging, aerospace, or any regulated supply chain, understanding AOQ is essential because it bridges a gap between incoming quality and delivered quality. Many teams monitor incoming defect percentages and sample acceptance rates, but fewer calculate the average outgoing quality. That is a missed opportunity because AOQ is the metric that ties inspection strategy directly to customer risk.

What average outgoing quality means

AOQ measures the average fraction defective that leaves your process when a rectifying inspection plan is used. Under rectifying inspection, accepted lots go forward with any uninspected defects still present, while rejected lots are typically subjected to 100 percent screening and defective units are corrected, replaced, or removed. Because of that, rejected lots contribute little or no outgoing defects, while accepted lots are the primary source of outgoing nonconformities.

The standard working equation is:

AOQ = P_a × p × (N – n) / N
where P_a is the probability of acceptance, p is the incoming fraction defective, N is lot size, and n is sample size.

This equation assumes that the sample itself is inspected and that rejected lots are rectified. The term (N – n) / N reflects the fact that some portion of the lot has already been inspected in the sample, reducing the number of potential defectives that can escape in accepted lots.

Why AOQ is different from AQL and AOQL

AOQ is often confused with AQL and AOQL. These terms are related but not interchangeable:

• AQL means acceptable quality level. It is a design point used when selecting a sampling plan, not a direct measure of outgoing quality.
• AOQ is the actual expected outgoing quality for a specific incoming defect level under a specific plan.
• AOQL means average outgoing quality limit, which is the maximum value of the AOQ curve. It represents the worst average outgoing quality that a rectifying sampling plan will produce.

In quality assurance meetings, this distinction matters. AQL helps you set the plan. AOQ tells you what the plan delivers. AOQL tells you the peak risk level of the plan across defect rates.

The variables you need before calculating AOQ

1. Lot size N: The total number of units in the batch submitted for inspection.
2. Sample size n: The number of units selected for inspection from the lot.
3. Incoming fraction defective p: The estimated percentage defective before the lot is accepted or rejected. In calculations, percentages are converted to decimals, so 1.5% becomes 0.015.
4. Acceptance number c: The maximum number of nonconforming units allowed in the sample for the lot to be accepted.
5. Probability of acceptance P_a: This may come from a binomial model, a Poisson approximation, or a known operating characteristic curve.

If you already know the probability of acceptance from internal software or validated statistical tables, you can calculate AOQ directly. If not, P_a can be estimated statistically based on your sampling plan.

How to compute probability of acceptance

For a single sampling plan, the probability of acceptance is the probability that the sample contains no more than c defectives. Two common methods are used:

• Binomial model: Best when each unit can reasonably be treated as defective or nondefective with a constant defect probability.
• Poisson approximation: Often used when the defect rate is low and the sample size is moderate or large, making binomial computations simpler to approximate.

For the binomial method, the acceptance probability is the sum of probabilities from 0 defects through c defects in the sample. For the Poisson approximation, the same logic applies, but the mean number of defectives in the sample is approximated by n × p.

As an example, if your lot size is 1,000, your sample size is 80, your incoming defect rate is 1.5%, and your acceptance number is 2, the acceptance probability under the binomial model is the probability of finding 0, 1, or 2 defectives in a sample of 80 when the underlying defect probability is 0.015. Once that probability is known, you multiply it by the incoming defect level and the finite lot correction term (N – n)/N.

Step by step example

Suppose you inspect consumer electronics assemblies with the following plan:

• Lot size N = 1,000 units
• Sample size n = 80 units
• Incoming defect rate p = 1.5% = 0.015
• Acceptance number c = 2

First, estimate the probability of acceptance. Using the binomial model, the probability of observing 0, 1, or 2 defective units in the sample is roughly 0.809 for this plan. Then calculate AOQ:

1. Compute the uninspected share of the lot: (1000 – 80) / 1000 = 0.92
2. Multiply acceptance probability by incoming defect fraction: 0.809 × 0.015 = 0.012135
3. Apply the finite lot term: 0.012135 × 0.92 = 0.0111642

The resulting AOQ is about 0.01116, or 1.116%. That means the average outgoing quality after the sampling plan and rectification is about 1.116% defective, equivalent to roughly 11.16 defectives per 1,000 units shipped on average.

Comparison table: incoming quality versus outgoing quality

The table below shows how outgoing quality changes when the incoming defect rate shifts under a fixed plan of N = 1,000, n = 80, and c = 2. These values are representative calculated examples that quality engineers often use to visualize AOQ behavior.

Incoming Defect Rate	Approx. Pa Binomial	AOQ	Defectives Per 10,000 Shipped
0.5%	98.55%	0.453%	45.3
1.0%	95.23%	0.876%	87.6
1.5%	80.89%	1.116%	111.6
2.0%	78.33%	1.441%	144.1
3.0%	56.83%	1.568%	156.8

The pattern is important: as incoming defectives increase, the lot is less likely to be accepted, which eventually prevents AOQ from increasing indefinitely. That is why an AOQ curve rises, peaks, and then falls. The highest point is the AOQL.

Benchmark table: sigma performance and defect opportunities

Many quality teams also connect AOQ thinking with broader process capability targets. The sigma-level benchmarks below are widely used in operations and continuous improvement discussions because they translate quality performance into defects per million opportunities.

Process Sigma Level	Approximate DPMO	Approximate Yield	Interpretation for AOQ Discussions
3 Sigma	66,807	93.32%	Too many defects for most critical outgoing quality targets
4 Sigma	6,210	99.38%	Good for many commercial applications, but not necessarily low enough for high-risk products
5 Sigma	233	99.9767%	Strong process capability with substantially lower outgoing defect risk
6 Sigma	3.4	99.99966%	World-class benchmark for extremely low defect environments

Although AOQ is not the same thing as process sigma, these benchmarks help explain why strong process capability lowers both incoming defect rates and the burden placed on acceptance sampling.

Common mistakes when calculating AOQ

• Using percentage values without converting to decimals. A defect rate of 2% must be entered as 0.02 in the formula.
• Ignoring P_a. AOQ is not just p × (N – n)/N. You must account for the probability the lot is accepted.
• Confusing accepted lot quality with outgoing average. AOQ is an average across accepted and rectified rejected lots.
• Applying the method when rejected lots are not rectified. If rejected lots are returned without screening, classic AOQ assumptions are not met.
• Using a rough approximation without validating it. Poisson can be helpful, but binomial may be better depending on defect rate and sample size.

How to interpret your AOQ result in practice

A low AOQ means your sampling plan and incoming quality together are producing a low expected outgoing defect rate. That is usually good, but interpretation still depends on product risk. An AOQ of 0.5% might be acceptable for low-cost packaging components and unacceptable for sterile medical items or aerospace fasteners.

Use AOQ with context:

1. Compare the result with customer quality requirements.
2. Check whether the resulting outgoing ppm aligns with warranty, service, and recall risk tolerance.
3. Review the sampling plan cost. A lower AOQ often requires either better incoming process quality or a stricter sampling plan.
4. Examine the AOQ curve rather than a single point estimate. This reveals where the plan performs worst.

When organizations report only incoming defect percentages, they can miss the customer-facing impact. AOQ turns abstract defect rates into a more realistic expectation of what leaves the facility.

When to use AOQ and when not to use it

AOQ is especially useful when:

• You use single sampling acceptance plans.
• Rejected lots are subject to complete screening or effective rectification.
• You want to estimate average quality delivered to customers.
• You are comparing alternative sample sizes and acceptance numbers.

AOQ is less suitable when:

• Rejected lots are scrapped or returned without complete screening.
• The inspection process itself introduces damage or misses a large share of defects.
• Lot quality is highly mixed or not well represented by a single defect fraction.
• You need defect opportunity metrics rather than unit-based defect metrics.

Authoritative references for further study

If you want to validate your methodology or go deeper into acceptance sampling theory, these sources are excellent starting points:

• NIST Engineering Statistics Handbook: Acceptance Sampling
• U.S. FDA Quality Systems Inspection Technique
• Penn State Department of Statistics: Sampling and Quality Control Resources

Final takeaway

To calculate average outgoing quality, you need more than the incoming defect rate. You need the structure of the sampling plan, especially the probability of acceptance and the share of the lot left uninspected when a lot is accepted. The most common practical equation is AOQ = P_a × p × (N – n) / N. Once you understand that relationship, you can compare plans, estimate outgoing risk, and justify inspection strategies with much greater precision.

Use the calculator above to test different lot sizes, sample sizes, defect rates, and acceptance numbers. Try varying the incoming defect rate to see how the AOQ curve changes. That visual pattern often reveals more than a single point estimate and can help you identify whether your current plan is truly aligned with your customer quality expectations.

This calculator is intended for educational and operational planning purposes. For regulated environments or contract-critical applications, validate the chosen statistical model and assumptions against your quality system requirements.