Just-In-Time: The Buffer Was Hiding Something

For most of industrial history, factories held large stocks of components. Parts arrived in bulk, sat in a warehouse, and were drawn down as needed. The reasoning was straightforward: running out stops production, and stopping production is expensive.

Just-in-time inverted that. Components arrive as they are needed — sometimes hours before use — and almost nothing sits in storage.

The immediate benefit is financial and substantial. Inventory is money already spent and not yet earned back, plus storage, insurance, handling, and the risk of it becoming obsolete. A factory holding weeks of stock has an enormous amount of capital sitting still.

That much is widely understood. The part usually missing from the summary is why the method's originators reduced inventory in the first place, which was not primarily about the cash.

Inventory as a place for problems to hide

The insight behind the original approach was that buffers conceal problems.

A supplier who is frequently late causes no visible disruption if you hold three weeks of their parts. A machine that produces defects at a low rate is absorbed by spare stock. A process that occasionally breaks down is invisible if there is enough work-in-progress to keep the next station busy.

None of these problems is solved by the buffer. They are hidden by it, and they continue costing money quietly.

Lowering inventory deliberately makes them surface immediately. The late supplier now stops the line. The defect rate now matters. The unreliable machine now has consequences that arrive within the hour rather than at the end of the quarter.

That was the point. Low inventory was used as a diagnostic instrument — a way of forcing problems into visibility so they could be fixed at the source. The reduced stock was as much a consequence of having fixed things as a cause of savings.

This is why the method came packaged with practices that get less attention: stopping the line when a defect appears, fixing root causes rather than symptoms, and building close relationships with a small number of highly reliable suppliers.

Cutting inventory without doing that second part gets you the savings and not the improvement — and leaves you with all the problems that were previously hidden, now unbuffered.

What the method assumes

Just-in-time works well under specific conditions, and it is worth naming them because they were true where it was developed and are frequently untrue elsewhere.

Supply is reliable. Deliveries arrive when promised, at consistent quality. Without this, no buffer means regular stoppages.

Supply is nearby. Short distances mean short lead times, which means problems can be corrected within hours rather than weeks.

Supply is responsive. A supplier who can adjust quantities quickly absorbs variation that would otherwise require stock.

Demand is reasonably predictable. Wild swings need buffering somewhere, and if it is not inventory it becomes overtime, expedited shipping, or missed orders.

Relationships are long-term. The original approach depended on deep supplier partnerships, not on continuously re-tendering to the cheapest bidder.

The difficulty is that as the method spread globally, it was frequently adopted as hold less inventory while the conditions were quietly abandoned. Supply chains stretched across oceans, sole-sourced to whoever was cheapest, on adversarial terms — and the buffers were removed anyway. The savings were booked continuously; the fragility appeared all at once.

Efficiency and resilience come from the same budget

The honest summary is that lean is a genuine achievement with a cost that only appears under stress, and both halves matter.

Inventory is a shock absorber. Removing it means removing the thing that would have absorbed a disruption. Where supply is reliable and disruptions are rare, that trade is strongly positive — decades of freed capital against occasional trouble. Where supply is fragile, it is a bet that keeps paying small dividends until it loses a large amount at once. This is the pattern described in black swan events: a long calm period justifies removing the buffers that exist for the events not yet in the record.

It also interacts badly with the bullwhip effect. Lean systems have no inventory anywhere to dampen demand signals, so small fluctuations amplify upstream more violently.

The workable position is not a choice between lean and stockpiling. It is buffering selectively: hold stock where lead times are long, sources are few, and the item is cheap relative to the cost of stopping. Run lean where supply is genuinely reliable and responsive.

And keep the part of the original method that gets dropped. Low inventory as a way of exposing problems is valuable. Low inventory as a way of reporting improved capital efficiency, while the problems it was meant to expose go unexamined, is how a diagnostic tool becomes a hidden liability.