A 3PL Weighs the CEO's Six Sigma Shopping List
A 240-person third-party logistics firm ships about 4,000 orders a month. On-time delivery has slipped to 91% against a 97% customer requirement, and the CFO estimates each late order costs $85 in expedites and credits. The CEO returned from a seminar and wants to fund 40 Green Belt seats ($180,000) plus a full-time Master Black Belt ($150,000). There is no continuous-improvement staff, no project pipeline, and order data sits in three separate IT systems. The operations manager, who controls the schedulers, is skeptical: "We already know the trucks leave late." The CFO wants a decision by Friday and has asked whether the program will pay for itself.
1. Using cost-of-poor-quality and project-pipeline logic, evaluate the CEO's proposal. What funding level and structure would you recommend, and what would you tell the CFO?
Read the worked answer
Size the problem first. 9% of 4,000 orders = 360 late orders/month x $85 = about $30,600/month, or roughly $367,000/year. The opportunity is real and large enough to justify a program.
Then test the proposal against capacity to use it. 40 Green Belts implies roughly 40 chartered projects. At 240 employees, with no CI staff, no pipeline and data split across three systems, the firm cannot scope or support 40 projects. Trained belts without projects produce no savings, so most of the $330,000 buys certificates.
Recommendation: a gated, staged start - one experienced coach (part-time MBB or external), 2-4 Green Belts on scoped projects, 8-12 Yellow Belts among shift supervisors, and one project on late departures. Fund the next tranche only after measured, Finance-validated savings.
Payback framing for the CFO: moving late orders from 9% to 5% recovers about $136,000/year. That easily covers a modest staged program but does not cover $330,000 in year one. Also flag the assumption: the $85 figure is an average and may not hold at higher lateness rates.
2. Three candidates are on the table: (A) cut late departures from 9% to 5%, (B) replace the warehouse management system, (C) reduce dock-worker turnover from 40% to 25%. Which should be chartered first, and what disqualifies the others for now?
Read the worked answer
Charter A. It passes the project-selection screen: a metric that already exists, a baseline (91%), a customer-linked CTQ (97% requirement), a clear link to COPQ, an effect achievable inside a few months, and the process largely inside the team's control.
B is a solution in search of a problem. Software is a countermeasure, not a problem statement, and there is no baseline showing what the current system costs. If the capital decision is already made, record it as a constraint in the charter and keep it off the problem statement.
C is a real issue but a poor first project. Long lag time before results appear, many confounders (pay, hiring market, supervision), and ownership sits in HR. Critically, turnover may be an input to A, so keep it as a suspected X to investigate inside A rather than a separate charter.
Also note the sequencing point: a project needs a named sponsor with resource authority, not just a team.
3. The operations manager insists the cause is obvious. How should the sponsor use him, and how does a Yellow Belt's role on this project differ from the Green Belt's?
Read the worked answer
Use the skeptic as an asset, not an obstacle. His statement "trucks leave late" is a hypothesis, not a finding. Make him process owner/SME on the team, give him the departure-time data, and let the team test whether scheduled vs. actual departure explains lateness. Skeptics who help generate the data usually accept the conclusion.
Role split:
- Sponsor/champion: owns the business case, clears barriers (getting IT to pull data from three systems), protects team time, validates savings with Finance.
- Green Belt: leads DMAIC, builds the data collection plan, runs the MSA and analysis, writes the control plan.
- Yellow Belt: works inside the process - collects and plots data, keeps run charts, applies local tools (standard work checks, 5S, waste walks), participates in FMEA, supports the countermeasure rollout. A Yellow Belt does not lead a full DMAIC project or run statistical tests alone.
Deploying Yellow Belts at the supervisor level is also the cheapest way to build the data habit this firm currently lacks.