Just-in-Time Inventory: What It Requires Before It Saves You Anything

Just-in-time inventory does not reduce the stock your business depends on. It moves that stock onto your supplier’s truck, and it only works if the truck is reliable enough to bet a production line on. That is the honest one-line summary of JIT management, and it is the part most explanations skip on the way to describing the savings.

What just-in-time actually means

Just-in-time inventory is a replenishment approach: order and receive stock as close as possible to the moment it is needed, so that little or nothing sits waiting. Applied to manufacturing it means components arriving near the point of consumption; applied to distribution it means goods arriving near the point of sale.

The just-in-time JIT system is often described as a philosophy, which is true and not very useful. Practically, it is a set of ordering rules that only function when the conditions underneath them hold. Inventory just in time is the outcome; the conditions are the subject worth your attention.

One distinction to clear up immediately, because the two get conflated constantly. Perpetual and periodic are recording methods — how stock movements reach your books. JIT is a replenishment method — when and how much you buy.

You choose one of each, and they are unrelated decisions except in one direction, covered below. If the recording question is the live one for you, perpetual versus periodic inventory is the article for that.

What JIT is actually trying to buy you

Four benefits are usually listed. They are not equally important, and pretending they are is how JIT gets sold into operations that should not run it.

Released working capital. This is the real motive for most firms, and it is worth saying so plainly. Stock is cash converted into a form you cannot spend. Holding four weeks of a component instead of one frees the difference for something else.

Less storage. Genuine, but usually second-order — unless you are physically out of space, in which case it briefly becomes the whole argument.

Less obsolescence. Matters enormously for anything dated, seasonal or subject to design change; matters very little for fasteners.

Faster detection of quality problems. The one nobody expects and the one operations people rate highest. If you hold twelve weeks of a component and it is defective, you find out twelve weeks late and you own all of it. Small, frequent deliveries surface the problem on the second batch.

Be clear which of these you are actually buying before you start. A JIT programme justified on storage savings in a business whose real problem is working capital will be measured against the wrong number and judged a failure when it succeeds.

The four preconditions

This is the spine of the article. Each of these is a test you can apply to your own operation this week, and JIT requires all four. Three of them are about your suppliers.

1. Lead times that are predictable, not merely short. Predictability matters more than speed. A supplier who reliably takes fourteen days is easier to run lean against than one who averages six but ranges from two to twenty.

The test: pull your last twenty purchase orders for a critical item and look at the spread between order and receipt, not the average. If you have never measured this, you are not ready to decide.

2. Suppliers who will hold stock for you and deliver in small lots. JIT moves the holding cost up the chain, and your supplier knows it.

Some will absorb it for a committed volume; many will price it back to you; a few simply cannot, because they are small or importing themselves. The test: ask your three main suppliers whether they would deliver weekly instead of monthly at the same unit price. Their answer is your answer.

3. Demand you can forecast within a usable band. Not perfectly — usably. If your monthly requirement swings by a factor of three with no visibility, no ordering rule can protect a line running on two days of cover. Project-driven businesses and those with a few large, lumpy customers usually fail this test and should stop here.

4. Stock records accurate enough to trigger on. JIT ordering fires automatically off a stock position. If that position is wrong, the system orders the wrong thing at the wrong time with total confidence. This is the one precondition entirely within your control, and it is where most implementations should start.

Where JIT fails in this market

Most writing on this topic comes from places where the supply chain broadly works. Here, the honest position is different, and this section is why the article exists.

Precondition one — predictable lead times — is the one that fails most often, and when it fails the other three do not matter.

Imported components with multi-week variance. A part with a nominal six-week lead time that in practice lands anywhere between five and eleven cannot be run on two weeks of cover. The variance is the number that matters, and it is the number nobody records.

Single-source suppliers with no alternate. JIT assumes a missed delivery is recoverable. With one qualified supplier for a critical part, a missed delivery stops the line, and the buffer stock you removed was the thing that used to absorb it.

Port and customs delay. Clearance times are outside your control and outside your supplier’s. Any plan that treats the arrival date as the ship date plus transit is a plan that has not cleared a container recently.

Currency exposure that makes buying early rational. This one cuts directly against JIT and it is entirely legitimate.

Where a currency is moving against you, buying forward is a hedge with a real return, and the finance director holding stock deliberately may be making a better decision than the operations manager trying to eliminate it. That tension should be resolved explicitly rather than by whoever has more influence.

None of this means JIT is wrong here. It means the preconditions have to be tested rather than assumed, and that in most operations they will hold for some items and not others — which is the argument for the hybrid below rather than for abandoning the idea.

The middle ground most firms should run

The practical recommendation, stated as such: run proper reorder points across the board, and apply JIT only to the SKUs that pass all four tests.

That means calculating reorder points from measured lead time and its variance rather than from habit, and setting safety stock as a deliberate decision about the cost of a stock-out against the cost of holding — a number that differs wildly between a fastener and a component that halts a line.

Then segment. Items with reliable local suppliers, steady consumption and short lead times are JIT candidates. Imported, single-sourced or lumpy-demand items get a buffer sized to their actual variance, and nobody apologises for it.

Good JIT inventory control is selective. A blanket policy applied to every SKU is not a lean operation, it is an untested assumption applied at scale, and the first import delay will demonstrate that.

JIT stock control works precisely where it is aimed and causes damage where it is not, so a just in time ordering system should be configured per item class, not per company.

This is also the version that survives a board conversation, because it can be justified item by item rather than as a philosophy.

What JIT demands of your systems

Here is the link back to recording method, and it runs one way only.

JIT is unrunnable on stock figures that are true at month end. An ordering rule that fires on a stock position needs that position to be right now, which means every receipt, issue and adjustment posting as it happens. A business on periodic records cannot run JIT; it can only run guesswork with a JIT label.

Beyond that, three capabilities do real work. Supplier lead time has to be tracked as actuals, not stored as the number the supplier quoted — the variance between the two is the input to every safety stock calculation you will make.

Purchase requisitions need to fire automatically from the stock position, because a rule that depends on somebody noticing is not a rule. And consumption has to be visible at the point it happens, which on a production line means issues booked against a works order rather than at the end of a shift.

Kanban, briefly

Kanban is a signalling method, not a synonym for JIT, and the two are used interchangeably often enough to be worth separating.

Kanban answers "how does the next station tell the previous one to send more" — historically a physical card returning down the line, now usually an electronic signal. It is one way of implementing pull-based replenishment. JIT is the objective; kanban is a mechanism for it.

Kanban inventory management works well for steady, repetitive consumption where the signal loop is short and reliable. It is a poor fit where demand is lumpy or where the replenishment loop crosses a port, because a card cannot express "and add four weeks of variance".

If reading this has turned the question from policy into tooling, the prior decision is which category of system you are actually buying — stock accuracy or physical execution inside the building. That comparison is in warehouse management system vs inventory software.

Everything above reduces to one requirement: an ordering rule can only be as good as the stock position it fires on, and the lead times it is calculated from have to be measured rather than quoted. That is what to look for in inventory planning software — real-time stock, supplier lead time held as actuals, and requisitions that raise themselves. It genuinely spans three areas: supplier performance and price variance sit in procurement software, while line-side consumption and works-order issues belong to a production management system. Where those three run on separate systems, the reconciliation between them is usually longer than the lead time you were trying to shorten, which is the practical case for cloud erp over a stack of point tools.