Quality and Inspection

Defect Rate and PPM Formula

Defect rate measures how many defects occur per unit inspected. PPM (parts per million) normalizes the number for supplier scorecards, quality benchmarks, and customer reporting.

Formula

Defect Rate = Defects / Units Inspected

Variables

Understanding the Defect Rate and PPM Formula

Defect rate is the cleanest single measure of how often your process produces something wrong. Dividing 18 defective units by 2,500 inspected gives 0.0072, or 0.72%, telling you roughly seven parts in a thousand fail. On the shop floor it drives accept/reject decisions, feeds SPC trend charts, and gates whether a lot ships. PPM simply scales the same fraction by a million (7,200 PPM) so tiny fractions become whole numbers that supplier scorecards and customers can compare across lines.

The inputs come from your inspection record: Defects is the count found in the sample, and Units Inspected is the sample size. The critical distinction is whether Defects counts flawed units or individual flaws, since one unit can carry several defects. Keep the sample representative and reasonably sized; 2,500 units gives a stable estimate, but a 50-unit sample turns a single reject into 20,000 PPM. Always state the sample size alongside the PPM so readers can judge confidence.

Interpret against your industry. Automotive suppliers routinely target under 50 PPM and top performers hit single digits, so 7,200 PPM would fail most customer scorecards. General industrial and short-run job shops tolerate several hundred to a few thousand PPM. If PPM spikes, break it down by defect type using a Pareto chart; usually two or three modes drive most of the count. Watch the trend over many lots rather than reacting to one sample.

Worked Example

An inspection of 2,500 units found 18 defective units.

  1. Defect rate = 18 / 2,500 = 0.0072 = 0.72%
  2. PPM = 0.0072 x 1,000,000 = 7,200 PPM

Result: 0.72% defect rate (7,200 PPM)

Common Mistake

Confusing defects with defective units. One defective unit can have multiple defects. If your count is defective units, PPM reports defective parts per million. If your count is individual defects, PPM reports defects per million units. Be explicit which metric you are reporting.

Frequently Asked Questions

How do you calculate defect rate and PPM?
Defect Rate = Defects / Units Inspected. Divide the count found by the sample size: 18 / 2,500 = 0.0072, or 0.72%. To get PPM, multiply the rate by 1,000,000: 0.0072 x 1,000,000 = 7,200 PPM. PPM turns small fractions into whole numbers so suppliers and customers can compare quality across parts and lines on scorecards.
What is the difference between defects and defective units?
A defective unit is a single part that fails, regardless of how many things are wrong with it. Defects counts every individual flaw, so one unit with three scratches is one defective unit but three defects. This changes your PPM: 18 defective units in 2,500 gives 7,200 PPM defective, while 18 defects on fewer units reports defects per million. State which you mean.
What is a good PPM defect rate benchmark?
Automotive and aerospace suppliers target under 50 PPM, with top performers under 10 PPM. General industrial manufacturing commonly runs a few hundred to a few thousand PPM. The 7,200 PPM in the example (0.72%) would fail most automotive scorecards but may be acceptable for a low-volume job shop. Judge against your customer's contractual PPM limit, not a universal number.
How do I convert a defect rate percentage to PPM?
Multiply the percentage by 10,000, or convert to a decimal and multiply by 1,000,000. A 0.72% defect rate becomes 0.72 x 10,000 = 7,200 PPM, or 0.0072 x 1,000,000 = 7,200 PPM. To go the other way, divide PPM by 10,000: 7,200 PPM / 10,000 = 0.72%. One percent always equals 10,000 PPM.
Why did my PPM jump when nothing changed on the line?
Small samples make PPM volatile. With 2,500 units, one extra reject moves you about 400 PPM; with 100 units, one reject swings PPM by 10,000. A single spike often reflects sample size or a one-off event, not a real process shift. Confirm with a larger sample or several consecutive lots, and Pareto the defect types before chasing a cause.
What sample size do I need for a reliable defect rate?
Enough that a single reject does not dominate the number. At 2,500 units, each defective part is about 400 PPM, giving a stable read on a 7,200 PPM process. For low-PPM parts, you may need tens of thousands of units to see any defects at all. Base the sample on your AQL sampling plan or the confidence level your customer requires.