1. Tight scope
One process, one owner, one measurable Y, 3–8 months. Lab turnaround, discharge time, one report — not "fix the hospital".
Coach: 12-question screening, STAR goal, Scope & Y tree, warning when the charter covers more than one process.
Why Green Belt projects stall
The method works. Projects stall in the same places.
Projects that succeed have a tight scope, clean data, the right test, a sponsor, savings validated by finance and a control plan built on roles and routines. Projects stall between Measure and verified results. Those are exactly the points where the Taktstone coach checks the Belt every day.
What the published data says
17%of 127 trained Green Belts finished the two projects needed to certify, in one company study1
> 80%certified where projects had structure and coaching2
Measure & Analyzeare where most stopped projects die, according to 201 LSS experts3
4 of the top 8failure causes are the wrong project or the wrong scope4
Nine lessons from 32 public projects
What separates finished projects from stalled ones
From published Green Belt-sized projects in US and European hospitals, factories, public agencies, banks, insurers and IT teams. Their results, not ours — sources below.
2. Show the spread
Lead times are skewed. Strong projects report the median and the spread, not just the average — one discharge project cut the SD from 129 to 37 minutes.
Coach: method picker suggests medians and non-parametric tests for skewed data; every result explained in plain language.
3. Data first
Almost no published project shows a measurement or data check. ERP timestamps and ticket backlogs needed cleaning before anyone could trust them.
Coach: Gage R&R — or a data-accuracy audit for time-stamped data — before analysis, and a warning if it comes after.
4. Don't stop at Improve
Projects stall between the fix and the proof: "result not yet verified", schedules that slip phase by phase, trainees who drop out.
Coach: tollgate audit, ≥ 20 stable points after the change before acceptance, daily next steps.
5. Savings finance signs
Many published savings are projected or cost avoidance, mixing hard and soft money, never validated.
Coach: finance-validator role and sign-off; the audit flags savings not validated by finance.
6. A sponsor who shows up
Projects tied to strategy with a reviewing sponsor survive. Public guidance even says: no sponsor, cancel.
Coach: free approver seats, server-stamped sign-offs and emailed sign-off requests at each tollgate.
7. Control by roles and routines
Gains held where people tracked their own numbers, a role owned the process, and boards were reviewed — not where people were "retrained".
Coach: control plan with an owner and a reaction plan, control charts after the change.
8. Fit your industry
Service and research teams quit when the training "feels made for manufacturing".
Coach: industry packs for IT, healthcare, finance and AI/ML; plain-language explanations.
9. Clear data access early
In Europe: works councils for staff-level data, GDPR and ethics review for patient data, and IT teams stretched across projects.
Coach: stakeholder register and communication plan; plan the data extract in Define.
Free self-check · runs in your browser · nothing is sent anywhere
Where could your project stall?
Nine questions, one per lesson. Answer for the project you are working on now.
Sources: 1 WPI study of a retail Green Belt programme · 2 Texas Children's Hospital · 3 Antony et al., IEEE Transactions on Engineering Management (2022) · 4 Minitab survey via iSixSigma. Lessons drawn from public case reports, for example hospital discharge (iSixSigma), purchase-order cycle time (iSixSigma), Crane Army (army.mil), Beacon Hospital Dublin (PMC), call-centre transfers (Pyzdek Institute) and NHS guidance. Public cases, no affiliation; these are other organisations' results, not Taktstone results and not typical outcomes.