Every plant has one step that limits how much it can ship. In a high-mix plant it hides well: routings vary, the mix changes week to week, and every department looks busy. Here is how to find it.
Why it’s hard to see
Utilisation reports make every machine look important, and the busiest-looking step is not always the one that limits output. In high-mix plants the load also moves with the order mix, so the step that hurt last month may not be the one hurting now. Start from evidence, not from reputation.
Step 1 — Look for where work waits
The bottleneck has a queue in front of it. Walk the floor at shift start and note where work is waiting and how old it is. A queue that is always there, at the same step, is the strongest signal you will get.
- Where do jobs sit longest between operations?
- Which step do supervisors expedite around every day?
- Where does overtime keep landing?
Step 2 — Check it against your data
Pull three exports from your business system: open orders, routings with standard hours, and work in process by operation. Then compare, for each work centre, the hours of work due in the next few weeks against the hours it can actually run.
- Load against capacity: the step with the most work due relative to its real available hours is your first candidate.
- Age of work in process: the oldest work tends to sit in front of the constraint.
- Routings against reality: where the floor’s path differs from the routing, trust the floor and fix the data.
Step 3 — Confirm it
Ask one question: if this step lost an hour, would the whole plant ship an hour less? If the honest answer is yes, you have found the constraint. If the hour could be made up elsewhere, keep looking.
At a high-mix metal fabricator, finishing was the bottleneck — but the bigger problem was what it spent its time on. Coating adhesion failures traced back to inconsistent pre-treatment, dimensional rejects traced back to fixturing that varied by operator, and jobs sat on hold because travelers moved without complete inspection records. Every part reworked went through the bottleneck twice. Fixing those causes at the source gave the bottleneck back capacity the plant had been paying for twice — without buying a machine.
At a powder-coating plant, the single largest consumer of line time wasn’t coating at all — it was colour changeovers that needed a full booth clean-out. Sequencing jobs by colour family and standardising racking so each pass carried more parts recovered real capacity on the existing line, without a second shift.
Step 4 — Decide what to do with it
Once it is named, the next moves follow the Theory of Constraints: get the most out of the bottleneck, make every other step follow its pace, and only then add capacity. That is what a schedule built around the bottleneck does every day.
Take the free scorecard — ten questions, instant result.
Take the scorecard