OperationsFramework / modelModelAccessible

Six Sigma level

How can six sigma level support strategic choice or positioning?

AccessibleStrategicTeam3 min read

Contents

A process-capability indicator that translates observed defects and opportunities into a sigma level.

The Six Sigma level indicates how consistently a process produces results within customer requirements. Motorola pioneered the metric in the late 1980s, and organisations including General Electric and Honeywell later promoted it widely. The customary goal is no more than 3.4 defects per million opportunities under the standard long-term-shift convention. To illustrate the implied rarity, imagine a goalkeeper playing 50 games in a season and facing 50 shots in every game: at the stated rate, one goal would be conceded roughly every 147 years.

When to use it

Answer the performance question: “How capable are our processes of delivering error-free work?”
Evaluate capability within the operational-processes and supply-chain perspective.
Define the data, formula, reporting cadence and sources required for a process-quality KPI.
Compare actual performance with an agreed target, benchmark or trend.

Origins

Motorola engineer Bill Smith is credited with introducing Six Sigma as a disciplined method for reducing defects and variation during the mid-eighties. Motorola connected process capability with defects per million opportunities, and General Electric’s later adoption under Jack Welch accelerated international use. The widely quoted defect threshold depends on a conventional long-term sigma shift, so any report should state its opportunity definition and statistical convention.

What it is

Perspective: Operational processes and supply chain perspective.

Key performance question: How capable are our processes of delivering error-free work?

Six Sigma is both a capability measure and an improvement methodology. As a methodology, it organises analysis and change through DMAIC:

Define internal or external customer requirements and the process outcome expected.
Measure current performance and establish how frequently defects occur.
Analyse the evidence and process map to locate causes, patterns and improvement opportunities.
Improve the selected process by designing and testing ways to remove, correct or prevent the verified problems.
Control the revised process so that it remains on the intended course.

Organisations commonly deliver DMAIC through trained internal practitioners—Master Black Belts, Black Belts and Green Belts—with responsibilities reflecting their experience and level of involvement.

The proposition is that a more capable process produces fewer customer failures and, when the project addresses a strategically important activity, contributes to stronger and more durable financial results.

How to use it

Measurement

Define the unit, defect and opportunity before collecting data. A sigma result is meaningful only when those definitions are stable and relevant to the customer.

Data collection method

Collect evidence from the process input, the activity itself and its output.

Input data describe what enters or initiates the process.
Process data describe efficiency and execution, including elapsed time, cost, value, labour and defects or errors.
Output data assess whether the resulting product or service meets its requirement.

Formula

A defect is any result outside the customer specification. An opportunity is one defined chance for such a defect to occur.

Calculate defects per million opportunities (DPMO), then convert that value to a sigma level using a declared conversion table.

The numerator is the total defects found. Units represent the outputs examined, and opportunities represent the defined ways in which each unit can be defective.

Frequency

Calculate the baseline at the beginning of the project, repeat the measure after improvement and continue at an appropriate interval during the DMAIC control stage.

Source of the data

Use available process records or generate observations through a controlled collection plan.

Cost/effort in collecting the data

Collection can be expensive when records are manual or definitions are inconsistent. Manual measurement is most defensible for strategically important processes. Automated operational systems can reduce effort, but their data still require validation.

Target setting/benchmarks

The conventional Six Sigma benchmark is 3.4 defects per million opportunities, so an organisation applying that convention would target 3.4 or fewer.

Example

A food-delivery team reviews 50 orders and records the following defects:

Delivery is late (13).

Food differs from the order (3).

Food is not fresh (0).

The resulting DPMO calculation is:

The stated yield-to-sigma table converts 106,666.7 defects per million opportunities to a sigma level between 2 and 3.

Sample levels of Sigma performance according to the Sigma conversion table

Top practical tip

Select Six Sigma projects from strategic priorities. A smaller number of material processes will usually create more value than numerous bottom-up projects aimed only at convenient, local savings.

Top pitfall

Do not confuse cost reduction with strategic performance. Some celebrated adopters reported major savings while still performing poorly in the market; process efficiency cannot compensate for weak strategic choices.