Most Plants Have One CI Program. The Reach Is What Varies.
Walk a well-run plant during shift change. The production cells have visual management boards refreshed every hour. Operators run a five-minute tier-1 meeting against takt time. Standard work is posted at every station. Andon cords get pulled on a clear set of triggers. The CI team is genuinely good at what they do.
Now walk to the planning department. The CI team's posters are on the wall. The improvement charter is signed. There's no daily management routine. The standard work for the master scheduler exists in someone's head. The visual control for backlog status is a spreadsheet that nobody pulls up in a meeting.
Same plant. Same CI program. Two different worlds.
One program, uneven reach
Every operating function in the plant has continuous improvement somewhere on its charter. Production and engineering of course, but also maintenance, planning, finance, IT, HR, EH&S, purchasing, materials, facilities. At many/most plants past a certain size, there's a single CI/Lean/OpEx team running a unified methodology: kaizen events, A3 templates, daily management routines, value-stream maps, the whole toolkit.
The methodology is genuinely strong, and the production cells where it's deeply embedded run well. The story isn't that there are separate programs to merge from every department. The story is what happens to one program once it leaves the cell.
Where the unevenness shows up
The CI program at most plants is genuinely strong on the production floor. It's the rest of the plant where the practice falls apart:
- Standard work is treated as an operator artifact. Engineers, planners, finance analysts rarely have documented standard work for their own processes. And when it exists, nobody updates it or audits to it.
- Visual controls populate the production cell. The front office runs on dashboards nobody opens and email status reports.
- Daily management routines are mature on the floor and ad hoc above it. The shift lead runs a five-minute tier; the planning manager runs an hour-long Tuesday meeting.
- Root cause analysis is rigorous when a quality escape happens. It's casual when a planning miss happens, an IT incident recurs, or a finance close goes sideways.
- Improvement events focus on production processes. The same A3 discipline rarely gets applied to a finance close cycle or an IT change-management process, even when those are the bottlenecks.
The gap is most visible at the boundaries, where production hands off to planning, where engineering hands off to maintenance, where finance hands off to operations. Standard work isn't agreed across the boundary. Visual controls don't span it. Daily routines don't connect.
What the data says is actually broken
When we look at the dominant root causes across the continuous improvement use cases in our library, the same handful surface repeatedly:
| Root cause | Use cases affected | Avg influence (1-5) |
|---|---|---|
| Incompatible System Formats | 10 | 3.9 |
| Missing Improvement Prioritization Framework | 9 | 4.3 |
| Improvement Gains Regression | 8 | 4.0 |
| Shared Improvement Ownership Without Accountability | 5 | 4.0 |
| Missing Long-term Capability KPIs | 5 | 3.9 |
| Analytical Method Application Gaps | 5 | 3.5 |
| Inconsistent Root Cause Analysis Methods | 4 | 4.2 |
Read these through the reach lens and they tell a coherent story:
- Incompatible System Formats. Central CI tooling is built for production data; finance, planning, and IT each have their own systems that don't aggregate into the CI tracker.
- Missing Improvement Prioritization Framework. The central team prioritizes against production goals; nobody prioritizes against support-function capability gaps.
- Improvement Gains Regression. Gains stick where the CI team has presence and decay where they don't.
- Shared Improvement Ownership Without Accountability. The central team owns the methodology, the line owns delivery, and nobody owns the fact that finance never adopted any of it.
- Missing Long-term Capability KPIs. Capability KPIs exist for production (OEE, takt adherence, scrap rate). Comparable KPIs don't exist for finance close cycles, planning accuracy, or IT incident recovery.
- Analytical Method Application Gaps. A3 doesn't translate cleanly to office contexts, and rather than adapting it, support functions stop trying.
This isn't a culture problem. It's a reach problem.
Why this gets misdiagnosed
The diagnosis depends on where you sit:
- The central CI team sees adoption resistance → tries to mandate broader rollout, deploys more change agents, escalates to leadership when functions aren't engaging.
- Support functions see CI methods built for production and conclude this isn't really for us → adopt the vocabulary without the practice.
- Plant leadership sees CI as a production-team activity → expects production-team results and isn't sure what good would look like in planning or finance.
- Operators experience CI as something they do. Support function staff experience it as something they're asked to attend.
Each of those is internally consistent. None of them name the actual issue: the program is methodologically sound but lacks the structural reach to make non-production functions practice it the same way.
What actually works
When we look at the most heavily-used enablers across those same use cases:
| Enabler | Use cases addressing | Type |
|---|---|---|
| Cross-Functional Governance Structures | 9 | Govern |
| Real-Time Operational Dashboards | 9 | Sense & Monitor |
| ROI Tracking and Sustained Investment Justification | 8 | Invest |
| Cross-Functional Planning and Review Cycles | 7 | Govern |
| Training in Structured Problem-Solving Methodologies | 7 | Train & Upskill |
| Adoption of Structured Problem-Solving Methods (A3, Kaizen) | 6 | Standardize |
| Cloud-Based Centralized Data Platform | 5 | Integrate & Connect |
Three patterns:
-
Cross-functional governance shows up twice in the top seven. Same finding as last week's accountability post. The central CI team can't extend reach alone. Cross-functional governance forces shared ownership of the practice, not just the charter. It's how you get a finance director to actually run a daily standup, not just say they support CI.
-
Real-time visibility, with sustained financial signal, and the plumbing to feed it. Dashboards built for production need to be built, with the same rigor, for support functions. ROI tracking has to follow improvements past their close gate, not stop at it. And centralized data platforms appear in this list because the #1 root cause, Incompatible System Formats, makes cross-functional visibility impossible without addressing the plumbing first. The "Improvement Gains Regression" cause exists because nobody's watching after the project closes.
-
Adoption, not just training, of structured methods. A3 and Kaizen training appear here, but lower than governance and visibility. The keyword in "Adoption of Structured Problem-Solving Methods" is adoption, which means actually practicing them in non-production contexts, with translation where the templates don't fit.
What's not on the list: more events on the production floor, more lean certifications for operators, more CI champions in production. Not because those are bad. The leverage isn't where the program is already strong.
The diagnostic question
If you're walking into a plant with a CI program that's plateaued, the diagnostic question isn't "do we need more events?" It's:
"For the functions in your plant outside the production floor, what does CI actually look like there?"
Specifically:
- Does your finance team have standard work for the close cycle? Does anyone audit to it?
- Does your planning team run a daily management routine that surfaces backlog risk before it hits the floor?
- Does your IT/OT team practice RCA on incidents the way Quality practices RCA on escapes?
- Does your HR team have visual controls for hiring and onboarding cycle time?
- Are improvements in support functions tracked over time, or are they one-time projects that quietly revert?
If most of those are "no" or "kind of," your CI program has a reach problem. Production has the methodology working at depth; the rest of the plant has it on the charter but not in the practice.
Where to start
Don't start with another floor kaizen. Start with one support function.
Pick the function whose unevenness costs you the most: the planning team whose forecast accuracy keeps the floor in firefighting mode, the finance team whose close cycle delays decisions, the IT/OT team whose change-management process blocks improvement deployment. Run the diagnostic above. Find which of the dominant root causes is actually present. Apply the corresponding enabler. Almost certainly cross-functional governance plus a visibility mechanism designed for that function's real work, not a copy of the production cell board.
The unglamorous truth about continuous improvement at most plants is that the methodology isn't broken. The reach is. And reach is solvable. Just not by repeating what already works on the floor.