Overproduction

Overproduction is the lean concept of making more than the next process actually wants right now. It sounds innocuous, but every lean practitioner from Ohno onward has called it the worst of the wastes. The reason is causation. Overproduction does not just exist as one entry on the 8-waste list; it produces several of the others as side effects.

"Overproduction is the only waste that breeds. Fix it once and three others shrink."

How overproduction works

The mechanism is straightforward. When a process produces faster or sooner than the downstream process consumes, the extra output has to go somewhere. It sits in a queue, on a rack, in a bin, on a pallet. That's excess inventory. The extra inventory has to be moved when the downstream process is finally ready, which is transportation. The downstream process now has more material to sort through to find the right part, which is motion. Defects that occurred in the overproduced batch are not discovered until much later, when there are now hundreds of them, which makes defects more expensive to fix.

So overproduction is not one waste. It is one decision that produces multiple wastes. The reverse is also true: stop overproducing and several other wastes drop on their own.

The underlying cause is almost always systemic, not individual. Operators overproduce because the system encourages it: long changeover times that get amortized over big batches, performance metrics that reward hours running, supervisors who want to "stay ahead of schedule," push-based MRP systems that release work to the floor on a forecast. The lean countermeasure is pull. When the only trigger to produce is a downstream signal, overproduction becomes physically impossible. The signal does not exist, so the production does not happen.

Where overproduction shows up on a small shop floor

Imagine a 25-person machine shop making three product families for two distributors. The owner notices that lead time has been creeping up despite no increase in order volume. Machines are running. Operators look busy. But finished goods inventory has grown from 3 weeks of demand to 8 weeks, and orders are arriving 2 weeks late.

A waste-walk would find the cause quickly. The setup time on the main CNC mill is 90 minutes, so the operator runs every job in lots of 200 to amortize the setup. The next station, deburr, can only run about 40 of those parts per shift. So 160 parts pile up after every CNC run. The deburr operator is constantly looking through the pile for the right job, because four different jobs are stacked together. Defects from one CNC run aren't caught until the 80th part because nobody opens the bins from previous runs.

The fix is not buying inventory software. The fix is reducing CNC setup with SMED so smaller lots are economical, then installing a kanban signal between CNC and deburr so CNC only makes more when deburr finishes a tray. Within a quarter, finished goods drops from 8 weeks to 3, lead time drops from 6 weeks to 2, and the operators are doing the same volume of real work with less effort.

Common mistakes with overproduction

• Buying inventory tracking software to manage the excess. Software measures the symptom. Pull signals stop the cause.
• Reducing batches at one station without stabilizing upstream supply. Smaller batches surface upstream variation instantly. Stabilize first, then shrink.
• Treating it as an operator behavior problem. Operators overproduce because the system rewards it. Change the metric and the signal, not the worker.
• Cutting overproduction by cutting capacity. Removing a machine reduces overproduction by reducing the shop's ability to make anything. Pull signals reduce overproduction without losing capacity for real demand.
• Stopping at finished goods. Most overproduction lives between operations, not at the loading dock. A shop with zero finished-goods inventory can still overproduce constantly inside the four walls.

Overproduction and related Lean tools

Overproduction is the most consequential of the 8 wastes (and the original 7). It is the parent waste of excess inventory, waiting, and several others. Its standard countermeasure is just-in-time production triggered by kanban pull signals. The metric that surfaces it most clearly is inventory-turns (the lower the turns, the more overproduction is happening).