Most Lean Six Sigma implementations do not fail during training. They fail in the six months after training, when the belts are certified, the banners are hung, and the floor quietly goes back to running the way it always ran. If you have watched that happen, this article is for you.
What follows is the implementation sequence that has held up across union and non-union plants, machine shops, and large production centers. It is not the certification-course version. It is the version you run when you own a real floor with real constraints.
Why most implementations stall
Three patterns show up almost every time an implementation fades:
- Training substituted for doing. The organization measures belts certified instead of problems solved. Knowledge without application decays in weeks. Don’t take that the wrong way, certifications have their merit, but I’ve seen people with “no belt” run circles around Master Black Belts when it comes to implementation and actually moving the needle. Someone with a certification AND the ability to execute on a project can be an unstoppable force!
- Projects picked for convenience. Early projects get chosen because a champion likes them, not because the value stream needs them. When the first wins do not matter to operators, the floor tunes out.
- No daily system underneath. Lean Six Sigma projects improve a process, but only a daily management system holds the gain. Without one, entropy wins every time.
The sequence that works
1. Start with one value stream, not the whole plant
Pick the value stream where pain is visible and the leader is willing. Resist the urge to launch everywhere at once. A working model in one area recruits the rest of the plant better than any kickoff meeting.
2. Map the current state before choosing projects
Walk the flow end to end and build a simple value stream map with the people who run it. The map does two jobs: it shows where the waste actually lives, and it turns project selection from opinion into evidence. Your first projects should fall out of the map, not out of a brainstorm.
3. Train inside the work
Teach DMAIC and the core lean tools against the live project, in short sessions, close to the floor. A two-hour module applied the same week beats a two-week course applied never. Supervisors and operators should learn together, because the improvement will belong to both of them.
4. Run the first project to a visible result
The first project needs to finish, and it needs to finish where people can see it. Changeover time on a bottleneck machine, defects at a pain-point station, a scheduling problem the crews complain about weekly. When operators see their own problem shrink, the method earns its credibility.
5. Standardize before you spread
Lock the gain into standard work, train it, and put it on the audit cadence before starting the next project. Improvements that live in a slide deck do not survive the next crew rotation.
6. Connect projects to daily management
This is the step most implementations skip. Every closed project should hand its key metric to the area’s daily board, where the tiered meeting watches it. Projects improve the process; the daily system defends it. If you want the full mechanics of that daily layer, that is exactly what Managing for Daily Improvement was written to install.
Common mistakes to design out early
- Belt inflation. Certifying people who have never closed a project dilutes the whole program. Tie every certification to a completed, audited improvement.
- Tool worship. A control chart on a process nobody owns is decoration. Choose the simplest tool that answers the question.
- Reporting theater. If project reviews are slide shows for executives, the floor will treat the program as overhead. Review at the process, with the data on the wall.
- Ignoring the middle. Supervisors make or break the system. If they are not trained, coached, and given time, no amount of black-belt talent will save it.
How to know it is actually working
Measure the program the way you would measure any process:
- Problems closed at the lowest possible level, trending up
- Gains still in place at the ninety-day audit, not just at project close
- Operators raising issues without being asked
- Fewer repeat problems reaching the plant manager’s desk
Lean Six Sigma works. It has worked for decades. The difference between the plants where it sticks and the plants where it fades is never the statistics. It is the discipline of implementation: one value stream, real problems, training in the work, gains locked into standard work, and a daily system that defends every inch of ground the projects win.