Improving Metrics: DPPM

Defective Parts Per Million (DPPM) is a trending quality metric among K-Form customers. DPPM is essentially another method of measuring nonconforming material levels. Although K-Form firmly believes measuring quality is a component of success, industry must examine what is truly measured and the effectiveness of the outcome. Several flawed assumptions hamper most DPPM implementations from providing realistic measurement.

DPPM measures the defective parts versus opportunities. The failure of DPPM occurs when the implementation poorly defines opportunity. The first flaw is quantitative. Some implementations misrepresent the quantity of opportunities. We will use the following example set for discussion.

Basic Implementation

Opportunity #1 Opportunity #2 Opportunity #3
Description Deliver 20 rivets. Deliver 2 parts Make Opportunity #2
Purchase Opp #1
Deliver Assembly
Problem 1 Rivet Missing 1 Hole Missing 1 Rivet Missing
Opportunities 20 2 1
Conformance % 95% 50% 0%

This thought experiment reveals a fundamental flaw when DPPM only accounts for top-level Bill of Material (BOM) items. K-Form sees a distinct trend where customers request full mechanical assemblies opposed to individual parts. This reduces customer-side integration costs, uses more supplier capability, and consolidates overhead costs (purchasing, quality, delivery). The only downside is the added complications to supplier quality metrics. Comprehensive implementations measure opportunities to reflect each item on the indented (hierarchical) BOM from components to assembly.

Alternate Implementation

Opportunity #1 Opportunity #2 Opportunity #3
Description Deliver 20 rivets. Deliver 2 parts Make Opportunity #2
Purchase Opp #1
Deliver Assembly
Problem 1 Rivet Missing 1 Hole Missing 1 Rivet Missing
Opportunities 20 2 22 (23 w/Assembly)
Conformance % 95% 50% 95.45% (or 95.65%)

The alternative implementation provides a more accurate representation of supplier efforts.

The second flaw is a qualitative issue. Most DPPM scoring methods do not account for complexity. K-Form manufactures precision parts. A high tolerance, 5-axis machined part may bear the same value as a single COTS screw. K-Form might laser cut a simple sheet metal cover and machine a chassis for the same customer. The two parts carry the same score weight despite complexity. DPPM routinely compares difficult prototype parts to mature, production parts. Although this may be a lesser concern than quantitative issues, secondary weighted DPPM factoring complexity can level scores across the supplier mix.

The third flaw is also a qualitative issue and perhaps a component of the second flaw. Textbook implementations determine acceptable DPPM levels by conducting rigorous evaluation of process capability. More often, K-Form sees blanket DPPM levels arbitrarily set without process knowledge. Is the metric still valid if the required process (or by extension Value Stream) is incapable of meeting the DPPM score? K-Form has seen DPPM requirements as low as 50 (99.99995%) for prototype/low-volume customers. One mistake in ten years will place the supplier in the under-performing category. The constant disapproval takes a psychological toll and promotes a negative relationship.

An issue of lesser concern is the development of hybrid metrics. The most common example is collecting late deliveries into DPPM. Late delivery measurements possess a host of other technical issues. Hybrid metrics, although convenient, cloud precision and compile any inaccuracies. Separate metrics simplify supplier analysis and corrective action.

K-Form builds collaborative relationships with customers to drive success. Objective measurement requires thoughtful development and detailed management reflecting complex supply chains. We welcome customers to discuss quality and supplier management metrics with the K-Form Quality Manager.