Lean implementation: what has to change in how the plant is managed

Beginning in one area rather than announcing a programme

Pick a single area and change it properly. The best candidate runs a product that repeats often enough to see a pattern, has a supervisor who is curious rather than merely compliant, and sits somewhere people walk past. Avoid starting on the resource that currently limits output, because every trial there costs shipments and the trial will be abandoned the first bad week. Accept that the area will borrow good people and that its output may dip while equipment is moved and methods are rewritten. A plant-wide announcement before anything works produces sceptics faster than any failure, because everyone has seen a launch before.

The management habits that have to change first

Ask what a manager will do differently on Monday. Usually it means standing at a board on the floor at a fixed time each morning, going to the machine rather than requesting a report, asking what the process did rather than who was at fault, and holding improvement time in the schedule when the plant is busy. It also means not breaking the rules the team has just been asked to follow: the sequence stays, the release cap stays, the extra batch is not authorised because the machine happens to be set. Every override is watched, remembered, and treated as the real policy.

Why the tool-led rollout stalls

The common pattern is a training programme, a set of audits, and a coordinator who carries the whole thing personally. It produces marked floors, boards with charts, and a scoring system that improves for a while. Then the coordinator is promoted or leaves, and within a couple of quarters the area reverts, because nothing in the operating routine ever depended on the work continuing. The deeper problem is usually the measurement system: if supervisors are still judged on machine utilisation and absorbed hours, the plant is paying them to build stock and to run long, which is precisely what the tools are meant to stop.

What has to be true before any of it holds

There is a floor below which lean techniques simply cannot be applied. Equipment has to run predictably enough that a planned sequence means something. Material has to arrive at the line in the right quantity, or operators will hoard. Someone has to be capable of writing down how a job is done. Above all, people need permission to stop and report a problem without an argument. Where those conditions are missing, the sensible sequence is to stabilise first: fix the chronic breakdowns, sort out line-side supply, get the basic records honest. Attempting flow on an unstable base produces stoppages that discredit the whole effort.

Telling real change from a tidier-looking floor

Audit scores, painted aisles and photograph boards prove that a campaign happened, not that the plant improved. Better evidence: elapsed time from order release to despatch is shorter for the same products, expediting has become rare, fewer schedule interventions are needed each week, problems surface within a shift instead of at month end, and operators are changing their own standards without asking permission. Watch also for improvements that simply moved a cost elsewhere, such as stock pushed back to a supplier or an operation subcontracted. If nothing in the financial or delivery picture has shifted after the first area matured, the change was cosmetic.

Frequently asked questions

Do we need an external consultant to start lean?

Not necessarily, but you do need someone who has watched a plant change before, because the early mistakes are predictable and expensive in credibility. The risk with outside help is that the consultant becomes the owner of the work and the plant learns to host workshops rather than run improvement. Structure any engagement so that your own supervisors lead sessions from early on, and make the deliverable a trained internal team rather than a set of reports.

Should we train everybody before we change anything?

Broad classroom training ahead of any visible change is one of the reliable ways to waste money. People forget the content, and the gap between the training and the first real project convinces them nothing will happen. Train the team doing the first area, in the area, on the problem in front of them. Wider training makes sense later, once there is something on site to point at and internal people who can teach from their own experience.

The plant is too busy for improvement work. What then?

Being busy is the usual condition, and waiting for a quiet period means waiting forever. The realistic answer is to reserve a small, protected block of time in a single area and defend it, accepting that output there may be slightly lower during the change. If that cannot be conceded, the honest conclusion is that management is not yet willing to pay for improvement, and it is better to say so than to start something that will be cancelled at the first shortage.