Scaling a Six Sigma Deployment That Cannot Validate Its Own Savings
Vertex Molding launched a Six Sigma program 18 months ago. Two Black Belts and twelve Green Belts were trained. Forty projects were chartered; nine are closed. Projects claim $2.1 million in savings, but Finance has validated only $640,000, mostly from two scrap-reduction projects. Belt time is nominally 20%, but production supervisors routinely pull Green Belts back to the line during high-volume weeks; actual project time averages 6%. Sponsors were assigned by department rather than by process, and three sponsors have changed since chartering. Projects were chosen mainly from a list of recent customer complaints, without a cost-of-poor-quality baseline or a link to the strategic goals. The CEO now wants to train twenty more Green Belts and launch a second wave of thirty projects. The deployment leader asks for a candid assessment before approving the expansion.
1. Diagnose the deployment. Which three weaknesses most directly explain the gap between claimed and validated savings, and through what mechanism does each operate?
Read the worked answer
The three weaknesses are linked. First, project selection based on complaint volume without a cost-of-poor-quality baseline lets teams charter work whose benefit is either small or not measurable in financial terms. That is why only $640,000 of $2.1 million is validated: soft benefits such as avoided overtime or customer goodwill do not survive Finance review. Second, nominal 20% belt time collapsing to 6% means projects receive roughly one-third of the planned effort. Teams skip data collection, jump from symptom to solution, and leave benefits unproven; projects stall at Analyze or Improve, which matches nine closures out of forty. Third, sponsor churn removes the person accountable for scope and benefits. Without a stable sponsor, tollgates become informal, scope expands, and no one owns the post-control handoff. Training more belts before fixing these mechanisms would scale the same failure pattern: more charters, the same low validation rate, and further credibility loss with Finance and operations.
2. Recommend a revised project selection and benefits-validation process. Specify the gates, the roles, and the evidence required at each gate.
Read the worked answer
Use a stage-gate portfolio process. Gate 0, concept: the sponsor names a process owner, a COPQ estimate with a baseline period, a strategic objective, and a belt with confirmed capacity; Finance signs the savings category, hard or soft, before charter approval. Gate 1, Define: charter, SIPOC, VOC/CTQ, and measurement plan reviewed by the sponsor and Finance; scope must fit the belt's available hours. Gate 2, Measure: MSA and baseline capability accepted; Finance confirms baseline cost and volume. Gate 3, Analyze: root causes evidenced with data; any scope change re-approved. Gate 4, Improve: pilot results and implementation cost; Finance re-validates the benefit calculation. Gate 5, Control: control plan, process-owner acceptance, and a 90-day sustainment check before savings are booked. The deployment leader owns the portfolio calendar; the sponsor owns the business case; Finance owns validation; the belt owns technical evidence. Publish a monthly funnel showing projects by gate and validated versus claimed savings. Cap new charters at the capacity implied by 20% of available belt hours, not at a target project count.
3. What should the deployment leader recommend about the CEO's proposal to train twenty more Green Belts and launch thirty projects?
Read the worked answer
Do not approve the expansion yet. First, stop the leakage: reassign or release the twelve current belts to protect at least 15–20% project time, and close or cancel the weakest open projects. Second, run a pilot wave of six to eight projects under the new gates with the existing belts. Measure validated savings and cycle time per project. If the pilot produces, for example, 70% or better Finance validation and on-time gate completion, then expand training in cohorts tied to confirmed project demand and sponsor capacity. Training without project demand creates trained belts with no meaningful work, which is worse than a smaller, credible program. The recommendation should be framed as sequencing, not rejection: fix selection, capacity, and sponsor accountability first, then scale.