Opens in a new tab

What Does a Lean Six Sigma Yellow Belt Do? A Worked Example of Reducing Order Errors

Share on:

A Lean Six Sigma Yellow Belt supports improvement projects and leads narrowly scoped investigations using basic analysis. Lean Six Sigma International’s programme covers measurement planning and root cause analysis within DMAIC. In an order-error project, the Yellow Belt investigates discrepancies between the request and the delivered material, then helps test a corrective action (programme details: leansixsigma.org).

The worked example below applies published warehouse research to a proposed Yellow Belt assignment. The research teams reported the results; the Yellow Belt responsibilities described here are our proposed use of their findings.

Set the Yellow Belt’s scope around one order handoff

A 2025 study in the Brazilian Journal of Operations & Production Management examined inventory and internal supply processes at a Brazilian dairy facility. Its DMAIC project addressed picking errors, described as “separation errors,” under a Black Belt project leader. The case supports a narrower Yellow Belt assignment within that wider project. The published result belongs to the full project team (study: bjopm.org.br).

For the proposed Yellow Belt assignment, the scope runs from picking materials to supplying production. A SIPOC maps the authorised request to the material delivered. It also identifies the production team receiving that material. The problem statement concerns discrepancies between the request and the delivery. Inventory accuracy remains a separate project outcome.

The sponsor can authorise investigation of the picking-to-production handoff while reserving wider system changes for the project leader.

Measure the difference between error counts and error rates

Figure 15 of the dairy study reports 47 separation errors in October 2023 and 11 in October 2024. The calculation is (47 – 11) ÷ 47 × 100 = 76.60%, matching the authors’ reported reduction in October error counts (study: bjopm.org.br).

DMAIC comparison Recorded separation errors
October 2023 47
October 2024 11
Source: Barbosa and colleagues, Brazilian Journal of Operations & Production Management (2025), Figure 15, available at bjopm.org.br. These figures are error counts.

The Yellow Belt still needs a denominator. Error counts alone cannot establish the change in the proportion of orders containing an error. Calculating a nonconforming proportion requires the number of affected units and the total number examined (statistical definition: itl.nist.gov).

The proposed measurement plan separates affected orders from individual discrepancies:

  • Orders containing an error: each affected order counted once.
  • Total errors: each recorded discrepancy counted under an agreed classification.
  • Orders examined: the denominator for the proportion containing an error.

An order with several discrepancies contributes once to the first measure and several times to the second. Separate fields preserve the distinction between a nonconforming unit and its individual defects (statistical guidance: itl.nist.gov). The Yellow Belt and project leader need to agree on that counting rule before comparing periods.

A Pareto chart selects the investigation priority

A Pareto chart ranks recorded categories by their contribution (method reference: leansixsigma.org). For this assignment, the Yellow Belt prepares the chart from verified discrepancy records and takes the leading category into a process review. The ranking sets the investigation priority. Establishing the cause requires observations of the work.

In a separate warehouse DMAIC study published in Heliyon in 2023, the cause analysis identified poor or missing stock labels and absent process instruction manuals (study: pmc.ncbi.nlm.nih.gov). Those findings give the order-error investigation specific conditions to examine. The team still needs evidence that either condition contributes to errors at its own handoff.

The proposed Yellow Belt assignment uses a gemba walk to follow a disputed material request through the handoff. An Ishikawa diagram organises suspected causes, with a supporting observation beside each entry. Any entry without evidence remains an open line of investigation.

Test the verification step before expanding it

The dairy project introduced a delivery-verification checklist and standardised operating procedures alongside other corrective actions. Because the team implemented these changes as a package, the published improvement cannot be attributed to the checklist alone (study: bjopm.org.br).

The Yellow Belt’s proposed role is to support pilot testing of the verification step within the agreed handoff. The trial needs a named person to compare the request with the material. Its discrepancy-recording procedure must also identify who resolves an error before release. During the trial, the Yellow Belt observes whether staff can follow the proposed standard work under operating conditions.

For the sponsor’s review, I would pair the order-error measure with handoff time as a balancing measure. Any reduction in discrepancies then sits alongside the time added by verification, giving the sponsor evidence to judge the operational cost of the check.

Give the process owner a response plan

The dairy project established weekly performance meetings and formalised procedures during Control (study: bjopm.org.br). For the proposed Yellow Belt assignment, a short control plan assigns responsibility for reviewing the order-error measure at a stated frequency. It also records the agreed response when performance deteriorates.

Before closing the assignment, I would ask the process owner to walk through the response to a newly detected picking discrepancy. If responsibility for containing the affected order remains unresolved, that control-plan action stays open.

Lean Six Sigma International’s Yellow Belt Lean Six Sigma programme teaches DMAIC for simple operational problems. The curriculum covers measurement planning and performance monitoring, with case-study exercises that apply those tools within the LSSx.0 framework (programme details: leansixsigma.org).

No results