Just-in-Time (JIT) Inventory: When It Works, When It Breaks, and the JIT vs JIC Tradeoff

Just-in-time (JIT) inventory is a replenishment strategy that targets holding as little stock as possible by ordering materials, components, or goods only when they are about to be needed. The goal is to receive inventory precisely when consumption begins — not before — eliminating the cost and risk of carrying buffer stock. In a closed-loop procurement platform — a system where every step from demand signal to purchase order to supplier reply to receiving runs in one connected record — JIT is not a binary policy but a per-SKU target, applied where demand and lead-time variability are low enough to support near-zero buffers, and replaced by statistical JIC buffers where they are not.

Quick answers

What is just-in-time inventory? JIT is a replenishment philosophy that minimizes on-hand inventory by receiving goods as close as possible to the moment they are needed. It originated in automotive manufacturing as part of Toyota's production system, where the goal was to expose process problems by eliminating the buffer inventory that would otherwise absorb and hide them. Applied to SMB purchasing, JIT means ordering frequently in small quantities from reliable suppliers rather than carrying large safety stock.

What is JIC inventory? JIC stands for Just-in-Case — the opposing strategy. A JIC approach intentionally carries buffer stock above expected demand to absorb variability in demand or supplier lead time. Safety stock is the formal quantification of that buffer: the extra units held to achieve a target service level when reality deviates from the forecast.

Does JIT work for SMBs? It depends on the SKU and the supplier. JIT works for high-velocity items with smooth, predictable demand and suppliers who deliver consistently. It breaks for items with intermittent, erratic, or lumpy demand patterns, or for any supplier whose lead time is long or variable. Most SMB catalogs need a hybrid — JIT-like tight buffers on predictable items, full statistical JIC buffers on volatile ones.

How is JIT different from the reorder point formula? The standard reorder point formula is ROP = (consumption rate × lead time) + safety stock. A pure JIT strategy is the limiting case where safety stock → 0, and ROP reduces to just the lead-time demand: ROP_JIT = consumption rate × lead time. That limit only holds when demand variance and lead-time variance are both near zero.

Origins

Just-in-time inventory was developed inside Toyota beginning in the 1950s and codified by Taiichi Ohno in his 1978 book, later published in English as Toyota Production System: Beyond Large-Scale Production (Ohno, 1988). Ohno identified inventory as one of seven forms of manufacturing waste (muda): excess stock hides quality problems, delays detection of defects, and ties up capital that could be deployed elsewhere. The solution was a pull-based production system where each workstation signals upstream only when it is ready to consume — carrying no buffer that would absorb and conceal problems.

The insight generalizes beyond automotive assembly: unnecessary inventory is a symptom of unreliable processes, not a cure for them. When supplier lead time is variable, the conventional response is to carry more buffer. Ohno's prescription was to fix the lead time variability instead. That prescription is harder at SMB scale, where buyers cannot dictate supplier operations the way Toyota could.

The math behind JIT

In continuous-review replenishment, the standard reorder point formula is:

ROP = (consumption rate × lead time) + safety stock

where safety stock = z × σ_d × √(lead time).

JIT implies driving safety stock toward zero, which requires two conditions to hold simultaneously:

• σ_d ≈ 0 — daily demand variance is negligible; the item sells predictably every period
• σ_LT ≈ 0 — lead-time variance is negligible; the supplier delivers in exactly the same number of days every cycle

Under both conditions, the reorder point collapses to:

ROP_JIT = consumption rate × lead time

Place the order exactly lead time days before stock depletes, and the shipment arrives just in time. No buffer, no carrying cost, no spoilage risk.

The catch: both conditions are idealizations that hold only for specific SKU-supplier combinations. For most real-world SMB purchasing, at least one variance is non-trivial.

JIT vs JIC

Neither strategy is universally correct. The practical question is which strategy applies to which SKU — and that question has a principled answer based on demand-pattern classification.

Why pure JIT breaks at SMB scale

Supplier lead-time variability. When lead time is not constant, the safety stock formula must account for both demand variance and supply variance:

safety stock = z × √(μ_LT × σ_d² + μ_d² × σ_LT²)

where σ_LT is the standard deviation of lead time across actual purchase orders. If your supplier quotes a 3-day lead time but delivers in 2–6 days in practice (σ_LT ≈ 1.4 days), the pure JIT zero-buffer target becomes untenable — you will stock out on the fast delivery cycles and over-receive on the slow ones. Supply variability alone can triple the required buffer relative to the naive formula.

Intermittent demand. For items that sell zero units on many days and a handful on others, a near-zero buffer means any demand coincidence produces an immediate stockout. The Syntetos–Boylan Approximation (SBA) is the bias-corrected method for forecasting these items, and it sizes a meaningful buffer precisely because demand intermittency makes zero-buffer impractical. An item classified as intermittent in the SBC framework — Average Demand Interval greater than 1.32 periods — should carry a JIC buffer, not a JIT near-zero buffer.

MOQs and pack sizes. Even when demand is perfectly predictable, supplier minimum order quantities force batch ordering. If daily demand is 6 units and the MOQ is 24, the order covers 4 days of demand regardless of intent. JIT logic works at the order frequency level; MOQs override it at the order size level.

Perishability and decay. Items with a material decay rate have waste embedded in any buffer. A JIC buffer for perishables costs more than the carrying-cost formula alone suggests, because some buffered units expire before sale. But a JIT near-zero buffer for perishables also implies frequent small orders with higher per-order freight cost. The correct answer for most perishable items is demand-adjusted PAR levels — not pure JIT.

Late supplier replies in open-loop procurement. In an open-loop buying process — where the PO is emailed and the operator hears nothing until the delivery arrives — a supplier reply changing an ETA from 3 days to 7 days surfaces too late for a corrective reorder. Any JIT implementation depends on prompt, accurate supplier communication. Closed-loop supplier reply monitoring surfaces ETA changes when the supplier sends them, not when the item fails to arrive.

The 2026 hybrid model: per-SKU JIT/JIC assignment

The right framework is per-SKU policy assignment, not a single all-in or all-out stance. The SBC demand classification — Syntetos–Boylan–Croston — provides the principled basis for this assignment by measuring two parameters per item: ADI (Average Demand Interval) and CV² (coefficient of variation squared):

Smooth items — those that sell consistently every day — are the candidates for JIT-like safety stock sizing. Even there, safety stock does not go to zero; it is set at a lower service level (75–85% instead of 95%+) because the consistent velocity means a single missed delivery is recoverable. That is the practical SMB interpretation of JIT: tight, formula-driven safety stock on predictable items, not a dangerous zero-buffer policy.

Tariff volatility and supply-chain disruptions since 2025 have pushed many SMB operators to reclassify items they previously treated as smooth demand into erratic or intermittent categories — demand spikes, supplier substitutions, and import delays created variance that historical σ estimates did not anticipate. The appropriate response is to revisit ADI and CV² monthly for supplier-exposed items and adjust service-level targets upward during periods of elevated uncertainty.

How closed-loop procurement operationalizes JIT/JIC by SKU

In a closed-loop procurement platform, the hybrid JIT/JIC model is not a quarterly spreadsheet exercise — it is a routine that runs from live operating data:

1. Demand classification runs on rolling sales. ADI and CV² are recomputed from POS sales or production consumption. Each item is assigned its current demand pattern: smooth, intermittent, erratic, or lumpy.

2. Safety stock is sized by classification. Smooth items receive a lower z target, reflecting their JIT-like predictability. Intermittent and erratic items receive the full z × σ × √(lead time) buffer using SBA-derived demand estimates where applicable.

3. Reorder points reflect actual lead-time distributions. Lead time is derived empirically from PO history — date dispatched to date received — not from a supplier's quoted estimate. Actual σ_LT feeds the safety stock calculation automatically.

4. Supplier replies update the operating model before the delivery date. When a supplier replies with a new ETA, the system surfaces that change while there is still time to reorder from a backup supplier or accept the delay consciously. JIT breaks silently in open-loop environments; closed-loop monitoring makes the break visible.

5. Receiving updates the next cycle. If a supplier delivers short or late, the empirical lead-time distribution and fill-rate record update automatically. Future safety stock sizing reflects what the supplier actually did, not what the onboarding form said.

The result is a per-item safety stock number that reflects actual demand variance and actual supplier reliability — updated continuously from live procurement data, not reset manually once a quarter.

When to revisit JIT/JIC assignments

Review the stocking strategy for an item when any of these change:

• Supplier lead time shifts materially — new carrier, port disruption, supplier location change
• Demand pattern shifts classification — velocity spikes, goes seasonal, or goes intermittent
• A tariff or trade policy change affects supplier reliability or cost
• The cost of a stockout changes — the item becomes higher-margin or operationally critical
• MOQ or pack-size negotiations change the minimum batch size

The underlying math is stable. The inputs are not. A quarterly review of demand patterns and supplier lead-time distributions is sufficient for most catalogs. Items in the erratic or lumpy categories may warrant monthly review.

Start a 90-day free trial at linenow.co to see per-SKU demand classification, safety stock sizing, and reorder points applied to your own POS and procurement data.

Related

• Safety Stock: How to Size It Statistically — the formula that quantifies the JIC buffer, including z-scores, worked examples, and review cadence
• Reorder Point (ROP) Formula — when to order, and how JIT affects the trigger calculation
• Lead Time: Definition, Formula, and How to Measure It Accurately — the time variable JIT tries to eliminate variance from; the core constraint on zero-buffer strategies
• Syntetos–Boylan Approximation (SBA) — bias-corrected demand forecasting for intermittent items that cannot support JIT-like buffers
• Coefficient of Variation (CV) and CV² — the demand volatility measure used to classify SKUs into smooth (JIT-eligible) vs erratic (JIC-required) categories
• Closed-Loop Procurement — the workflow that keeps JIT/JIC assignments and safety stock numbers current from live demand and supplier data