Parts per Million (PPM)The defect rate as a key figure in supplier management
Procurement glossary · Updated · 9 min read
Also known as: PPM, Defects per Million, Defect rate in PPM, DPMO, PPM rate
Short answer
- PPM (Parts per Million) indicates how many defective units occur per one million delivered parts.
- The calculation is carried out using the formula: defective parts divided by total quantity times 1,000,000.
- As benchmarks, values below 50 PPM are considered good and below 10 PPM very good in series production.
Parts per Million (PPM) describes how many defective units statistically occur per one million delivered parts and is regarded as the standard key figure for delivery quality in industrial partnerships. The metric makes defect rates comparable that would be barely representable as percentages, and forms the basis for supplier evaluations, escalation processes and quality objectives in procurement contracts. Buyers use PPM to manage suppliers objectively, demand improvement measures and protect their own production from scrap and stoppages. The following article explains the definition, calculation, target values, typical pitfalls and the strategic significance of PPM in procurement.
Key facts
- PPM indicates how many defective parts occur per one million delivered units.
- The formula is: number of defective parts divided by total quantity, multiplied by 1,000,000.
- In practice, values below 50 PPM for series parts are frequently considered good and below 10 PPM very good delivery quality.
- PPM is typically evaluated monthly or on a rolling basis per supplier and material class.
- The key figure is usually fed by incoming goods inspection, complaint data and customer complaints.
- Exceeding agreed limits frequently triggers escalation levels such as containment or process audit.
Contents
- Definition: Parts per Million (PPM)
- Calculation and data basis: how PPM is determined in procurement
- Key KPIs around PPM and typical target values
- Methods and approach: managing PPM in supplier management
- Risks and controls: sources of error in PPM evaluations
- Current developments: PPM in the age of data integration and resilience
- Typical metrics
- Practical example
- Conclusion
- Frequently asked questions
Definition: Parts per Million (PPM)
Parts per Million (PPM) is a relative key figure for defect frequency, indicating how many defective units statistically occur per one million delivered parts. It originally comes from statistical process control and has established itself as an industry-wide, universally understood language between customers and suppliers. Instead of percentages with many decimal places, PPM allows vivid communication of very small defect rates.
Distinction from other quality key figures
PPM measures the number of defective units, while key figures such as on-time delivery or service level reflect delivery reliability. PPM differs from First Pass Yield and the scrap rate in one's own plant in that it looks at the quality of purchased parts. In many companies, several key figures are combined into a supplier rating model, in which PPM usually receives the highest individual weight in the quality category. Anyone considering PPM in isolation misjudges risks such as unstable processes, which only later manifest themselves in complaints.
PPM vs. DPMO and Cp/Cpk
Closely related is the key figure DPMO (Defects per Million Opportunities), which is rated per defect opportunity and is widespread above all in process quality. PPM, by contrast, refers to delivered units and is therefore easier to capture for supplier management. Process capability indices such as Cp/Cpk, in turn, are predictive: they describe a process's capability before defects occur. In practice, both perspectives complement each other – PPM shows the defect reality that has occurred, process capabilities the underlying risk profile.
Calculation and data basis: how PPM is determined in procurement
The calculation is formally simple, but in practice dependent on data quality. What is decisive is which defect events are included and at which point in the value chain they are captured. Without a clear definition, disputes arise over the validity of figures – for example between incoming goods inspection and the complaints department.
The calculation formula
PPM = (number of defective units ÷ total quantity of delivered units) × 1,000,000. If 1,200 parts are delivered and 3 of them are rejected, this results in 2,500 PPM. Typically, not the rolling monthly value but a rolling three- or six-month average is used to smooth out random effects with small lot sizes. Small delivery quantities strongly distort PPM: a single complaint on 100 delivered parts yields 10,000 PPM, although it is a single event.
Which defects are included in the key figure?
As a rule, the following count as defective: parts with a demonstrable material or manufacturing defect, documentation deficiencies such as missing inspection certificates, and packaging and labelling errors where these have been agreed. Shortages (that is a delivery reliability issue) and transport damage usually do not count, insofar as the transfer of risk under the purchasing terms and conditions or Incoterms such as CIF or DPU lies with the supplier or the carrier. It is important to agree the counting logic in writing, for example in the purchasing terms and conditions or in the quality assurance agreement.
Data capture: incoming goods, complaints, customer defects
PPM can be captured at three levels: incoming goods inspection, internal production reporting and end-customer complaint. Many companies evaluate levels one and two together and track customer complaints separately, as they cause the greatest damage. Clean master data management is decisive for meaningful figures: without Master Data Governance, faulty reports cannot be clearly assigned to supplier, article and batch. Batch and lot traceability is the prerequisite for reliable PPM evaluations.
Key KPIs around PPM and typical target values
PPM is only meaningful if it is embedded in a system of key figures. Target values depend heavily on industry, material class and technology – there is no rule of thumb for all parts. What is decisive is that target values are agreed contractually or as target agreements and are regularly monitored in procurement controlling.
Typical target values by industry context
In series production of safety-critical parts (for example automotive), values below 10 PPM are common, and sometimes zero defects is tolerated. In general mechanical engineering and electronics, target ranges between 50 and 500 PPM are frequently applied depending on part complexity. For custom fabrication or small series, stricter values are often unrealistic; here manual inspection processes with limited statistical validity dominate. The target agreement should be oriented to the technology, not to a blanket benchmark.
Complementary key figures in the supplier rating
PPM should always be evaluated together with further quality key figures: complaint lead time, 8D processing time, on-time and in-full delivery as well as audit grades. In the supplier rating model, these dimensions are typically weighted together, with the quality score often accounting for 30 to 50 per cent of the overall scorecard in practice. In addition, key figures on risk mitigation are relevant, such as the proportion of qualified alternative suppliers per material class as a result of active Supply Base Optimization.
Measurement and reporting rhythm
In practice, monthly evaluations per supplier and article group with a rolling twelve-month trend have proven worthwhile. The evaluation should be carried out separately by material class so that structurally problematic technologies do not disappear in the average. A plan-actual comparison of stocks of complaint items and open dimensional issues complements the picture. Regular review meetings (supplier evaluation discussions) maintain joint improvement pressure.
Methods and approach: managing PPM in supplier management
PPM is only valuable if exceedances are responded to appropriately. A staged response model, agreed in quality assurance agreements or as part of the terms and conditions, has proven its worth. It defines from which PPM value which measures take effect.
Stage model for limit exceedances
Typical escalation levels are: 1) written warning with a request for root cause analysis (8D report), 2) mandatory containment measure with 100 per cent end-of-line inspection by the supplier and labelling of the remaining quantity, 3) process audit at the supplier's site, 4) process redesign and renewed supplier qualification. Only in the event of repeated or serious violations does demotion or removal from the supplier master follow. It is important that the levels are agreed in writing and comprehensibly so that they are also legally defensible.
Containment and blocked stock management
Containment means the short-term sealing off of the defect path: the supplier inspects 100 per cent of its deliveries, the customer carries out intensified checks at incoming goods and affected stocks are blocked. Containment measures cause considerable additional costs in practice, which are frequently settled via a credit note or special agreement. The duration should be limited; permanent containment is a sign that root cause elimination at the supplier is stalling.
Integration into procurement strategy and supplier selection
Historical PPM values are a central criterion in supplier selection and in ongoing Commodity Strategy. For A-parts with a high risk of failure, dual sourcing is worthwhile in order to secure supply in the event of defects via a Business Continuity Plan (BCP). In the invitation to tender (request for quotation created with AI), PPM should stand as an evaluated criterion alongside price – a lower purchase price can quickly be eaten up by a high defect rate through scrap, rework and downtime. Quantity scales and payment terms can also be linked to quality performance, for example through rewarded discount tiers for stable quality.
Risks and controls: sources of error in PPM evaluations
PPM can be misleading if capture, attribution or target values are poorly defined. The most common sources of error lie less in the mathematics than in the organisation surrounding the key figure.
Small-quantity and selection effects
With small lot sizes, PPM is statistically hardly reliable – a single complaint produces astronomical values. In practice, this is solved by minimum quantities per evaluation period or by combining it with absolute complaint figures. A further effect arises from selection: anyone who only counts inspected parts implicitly assumes that uninspected goods are defect-free. Clear inspection plans with documented sample sizes are therefore a prerequisite for meaningful figures.
Delayed and omitted reports
A significant proportion of defects is not discovered in practice in the same month in which delivery took place. Without clear reporting deadlines and traceability via batch number, defects shift to later periods or disappear into unresolved reports. In addition, contractual limitation periods pose a risk: anyone who reports defects too late may lose claims – the respective purchasing terms and conditions and statutory limitation periods must be examined on a case-by-case basis.
Metric gaming and conflicts of objectives
Suppliers that are geared towards PPM targets occasionally try to remove defects from the key figure via blocked-part lists, interpretations of customer specifications or goodwill arrangements. Buyers should therefore regularly reconcile complaint shares with incoming goods and production data and audit critical figures on a sample basis. A further conflict of objectives arises when procurement optimises only on price: anyone who buys purely focused on purchase price typically pays for the resulting defects twice over later.
Current developments: PPM in the age of data integration and resilience
PPM remains a proven key figure, but the way it is handled is changing: real-time data, supplier portals and Predictive Quality are shifting the perspective from reacting to defects towards predicting them. At the same time, securing the supply chain is moving more strongly into focus.
Automated quality data and EDI interfaces
More and more companies are exchanging quality data automatically via supplier portals or EDI interfaces: defect codes, finding quantities and 8D status flow directly into the ERP. This shortens response times and reduces coordination effort. Clean Master Data Governance remains a prerequisite so that supplier and article identifiers can be clearly assigned across system boundaries. A standardised supplier identifier is the basis for this.
From PPM to Predictive Quality and process capability
Instead of only counting defects retrospectively, advanced companies rely on predictive approaches: process capability indices, supplier SPC data and machine feedback are combined to detect critical trends before they become supply problems. PPM remains the currency for communication with the supplier, while the prediction is carried out internally from process data. For buyers, this means: quality requirements are increasingly defined earlier and more bindingly in the project procurement phase.
PPM in sustainability and resilience contexts
The concept of quality is broadening: non-conformities in sustainability evidence, such as a missing FSC/PEFC certificate for wood components or missing material declarations, are also captured in extended PPM-like deviation rates. Increasingly, such key figures flow into a combined risk and sustainability rating. This shifts PPM from a pure production key figure to an overarching compliance indicator.
Typical metrics
| Metric | Benchmark | Context |
|---|---|---|
| Supplier PPM | typically < 50 PPM | Benchmark for series parts; safety-critical parts < 10 PPM |
| 8D processing time | often < 30 days | Time until the effective permanent corrective action following a complaint |
| Proportion of suppliers within target range | typically > 90 % | Proportion of active suppliers below the agreed PPM limit |
| Complaint lead time | typically < 14 days | Time from defect report to the supplier's statement |
Typical practical benchmarks – they vary by industry, part and contract.
Practical example
A mid-sized mechanical engineering company sources pressed parts for gearbox components from several suppliers. After repeated dimensional deviations from one supplier, that supplier's rolling six-month PPM value rises from around 40 to over 180 PPM. The defined limit of 100 PPM in the quality assurance agreement is exceeded, whereupon the agreed escalation cascade takes effect. Only after a process audit and the obligation to perform 100 per cent end-of-line testing does the value fall back below the target value within four months.
- PPM limits and escalation levels are agreed in writing in the quality assurance agreement.
- If the limit is exceeded, containment with 100 per cent end-of-line inspection at the supplier is initiated immediately.
- Batch traceability enables each defect to be assigned to a lot, supplier and process level.
- In the rolling supplier rating, PPM contributes a weight of around 40 per cent to the quality block.
Conclusion
Parts per Million (PPM) is more than a technical key figure: it translates quality into a currency that is binding for both parties and creates the basis for fair supplier evaluations and comprehensible escalations. The prerequisites are a cleanly agreed counting logic, reliable batch traceability and a staged response model that not only sanctions but also triggers improvement. Target values should be oriented to industry, material class and technology, not to blanket ideals. Combined with process capability data and automated data capture, PPM is increasingly transforming from a mirror of the past into an early warning indicator. Anyone who consistently anchors PPM in procurement strategy and supplier development demonstrably reduces scrap, downtime risks and complaint costs.
Frequently asked questions
What does PPM mean in procurement?
PPM stands for Parts per Million and describes how many defective units statistically occur per one million delivered parts. The key figure serves as an industry-wide, universally understood standard for assessing delivery quality. It is typically evaluated monthly or on a rolling basis per supplier and material class.
How is PPM calculated?
The formula is: number of defective parts divided by the total quantity of delivered parts, multiplied by 1,000,000. If five out of 10,000 delivered parts are rejected, this results in 500 PPM. It is important to clearly define which types of defects are included and from what point in time a report counts.
Which PPM values are good?
It depends on the industry and the part. In series production of safety-critical parts, values below 10 PPM are common, while in general mechanical engineering target ranges of 50 to 500 PPM are frequent. For small series and custom fabrication, such values are often statistically unreliable, as small quantities strongly distort the key figure.
What happens if the PPM limit is exceeded?
Typically, an agreed escalation cascade applies: first a root cause analysis via 8D report, then containment with 100 per cent end-of-line inspection, if necessary a process audit and finally process requalification or re-selection of the supplier. The steps should be fixed in writing in the quality assurance agreement. Additional costs arising from containment are frequently settled via credit notes.
Which defects count towards the PPM key figure?
Typically counted are material and manufacturing defects, documentation deficiencies such as missing inspection certificates, and agreed packaging and labelling errors. Shortages (delivery reliability) and transport damage generally do not count, unless the transfer of risk under Incoterms such as CIF or DPU lies with the supplier. The counting logic should be defined contractually.
How can PPM be manipulated?
Possible distortions arise from omitted or late defect reports, selection effects in incoming goods inspection and the reclassification of defects as goodwill cases. Buyers should regularly reconcile PPM figures with production and customer complaint data and audit critical cases. Sound Master Data Governance is a prerequisite for reliable evaluations.
