Four inventory replenishment methods compared (and which fits your catalog)

Every inventory replenishment method is a rule for answering the same two questions — when to order, and how much — and there are four that cover almost every small merchant: reorder point (a calculated trigger per SKU), min-max (two fixed thresholds), periodic review (a fixed calendar and an order-up-to level), and demand-driven (time-phased buffers sized to decouple lead time from demand). The right choice is dictated by your lead time, your demand stability, and how many thresholds a human can actually maintain — not by which method sounds most sophisticated.

Here's each rule in operational terms, the comparison table, and the failure mode that actually shows up in real stores.

Reorder point

The rule. Compute a per-SKU trigger: (daily velocity × lead time) + safety stock. When available stock hits the trigger, place an order — typically sized by an order-up-to calculation. Between orders, stock is monitored continuously (in practice: every day someone or something compares the number to the trigger).

Where it's strong. Long lead times with variable demand — the overseas-container case. The trigger moves with each SKU's actual velocity and actual lead time, so a fast mover and a slow mover are treated differently without anyone making a judgment call per item. It's the most defensible method per SKU, and it's the one every credible inventory tool computes under the hood. The formula and worked example are in the reorder point guide.

Where it fails. Input honesty and monitoring discipline. A trigger built on a quoted lead time sits 10 days late; one built on calendar-day velocity sits another 5 days late; and one that nothing actually compares against daily is decoration. The method's precision is also its maintenance cost: two inputs per SKU per quarter, times 400 SKUs, is real work.

Min-max

The rule. Two fixed numbers per SKU: a min that triggers, and a max you order up to. The order quantity is automatic — max minus current stock — so the buyer never decides quantity, only acts when the min is hit.

Where it's strong. Large catalogs of steady sellers, where ordering becomes a mechanical act and human attention is freed for the SKUs that deserve it. It's also easy to explain to a non-planner, which matters: a system nobody on staff understands gets bypassed.

Where it fails. Static thresholds meet changing demand. Min-max has no mechanism for the fact that a SKU's velocity doubles in Q4 or that its lead time stretched when the supplier changed factories. Worse, the min is usually set once from intuition and never revisited — so it encodes last year's business. And when the same min-max logic is applied uniformly across a catalog, a hero SKU and a dud get identical treatment, which is precisely backwards. If you use min-max, differentiate the thresholds by velocity band and put a review date on every SKU.

Periodic review

The rule. The calendar is the trigger. On a fixed cadence — every Monday, the first of the month, the day before your consolidated container — you count or read stock for the SKUs in scope and order each one up to a target level.

Where it's strong. Suppliers with order minimums or consolidated freight, merchants who batch buying into one admin block, and stores where someone counts weekly anyway. It also has a hidden benefit: reviewing the whole catalog in one pass surfaces cross-SKU patterns — a supplier drifting, a category slowing — that per-SKU triggers hide.

Where it fails. Inter-review gaps. If you order every two weeks, stock has to survive the lead time plus up to two weeks before the next chance to order. The target level therefore has to cover lead time + review period + buffer, and most stores set it to lead time + a bit. That's the standard failure: a correctly calculated target that forgot the review interval, producing a predictable shortfall in the days before each review.

Demand-driven

The rule. Instead of forecasting, place strategic buffers at decoupling points — typically finished goods you buy finished, or components before a long lead-time step — and size each buffer from average daily usage and lead time, split into zones. Replenishment is triggered by buffer penetration: when stock drops into the lower zone, order enough to refill the top zone; when it's deep in the lower zone, order urgently.

Where it's strong. Long, multi-stage supply chains with lumpy demand and unreliable lead times — the situation where forecasts are least useful and where the buffer-zone framing forces a useful conversation about which SKUs get which treatment. The "order to refill the buffer" logic is also immune to the forecasting failure mode of chasing a bad prediction.

Where it fails. Complexity and inputs. Buffer-zone sizing has several parameters per SKU, and a store with 300 SKUs and one supplier doesn't need the machinery — it needs a reorder point. And as with any method, a demand-driven system built on calendar-day averages is a sophisticated way of being wrong.

The comparison table

Picking per segment, not per store

The decision that matters isn't "which method" — it's "which method for which part of the catalog." A mixed policy is normal and correct:

• Overseas container SKUs with real cash per PO → reorder point with a measured lead time. The precision pays for the maintenance.
• The long tail of steady, cheap items → min-max with velocity-banded thresholds. Nobody should be doing statistics on a $6 accessory.
• Anything from a supplier with a monthly minimum or consolidated freight → periodic review, with the target set to lead time + review period + buffer.
• A multi-stage product where a long lead-time component is assembled late → demand-driven buffer at the decoupling point, if you have the volume to justify the parameters.

Two rules that apply across all four:

Set the review cadence before the thresholds. A trigger nobody reads is not a trigger. If your reality is a Monday buying block, then periodic review is your actual method whether or not you call it that — and the other methods have to be shoehorned into that cadence to work.

Store the inputs, not just the outputs. Every method here consumes the same two quantities: units per day and days of lead time. Keep those two per SKU, dated, in one place. Then changing methods is a recalculation, not a rebuild — and when you audit why a SKU went dark, you can see which input was wrong rather than guessing.

The input that breaks all four

There's one shared dependency worth naming on its own: velocity. All four methods consume units-per-day, and the standard way of computing it — units sold ÷ days in the period — treats every day a SKU was out of stock as a day of zero demand. A SKU that sells 7 a day and was dark 11 of the last 90 days measures at 6.1/day, a 12% understatement, and the understatement is largest on exactly the SKUs that go dark most often.

The consequence is the same in every method: a trigger, threshold, target, or buffer set too low, which causes the next stockout, which further depresses the measured rate. Whether you run min-max or demand-driven buffers, the loop is identical. The fix is to divide by days available rather than calendar days, and to sanity-check each SKU's resulting days of cover against its lead time — a quick cross-check that catches the failure before the shelf does.

A note on tooling: ClearStock, our live Shopify app, computes velocity and days of cover (on calendar days inside its review window — it does not exclude out-of-stock days) and classifies products as moving, slowing, stuck, dead, or oversold — useful for deciding which segment a SKU belongs in. It doesn't forecast or create purchase orders, so the ordering rule stays yours.

Common questions

Which replenishment method is best for a small store? The one that matches your lead time. Under a week with steady demand: min-max or par levels. Two to six weeks: periodic review or min-max with banded thresholds. Six weeks or more, variable demand, meaningful cash per PO: reorder point with a measured lead time. Most catalogs of a few hundred SKUs end up running two or three methods by segment.

What's the difference between min-max and reorder point? Both trigger on a stock level; they differ in where the quantity comes from. Min-max fixes both numbers in advance, so the order size is max minus current stock. A reorder point calculates the trigger from velocity and lead time, and the quantity from a separate order-up-to calculation — more accurate, more maintenance.

What's the difference between periodic review and continuous review? Continuous review checks stock constantly and orders the moment a trigger is crossed. Periodic review checks on a schedule and orders enough to reach a target — which must also cover the days until the next review, so lead time + review period + buffer.

Is demand-driven replenishment worth it for a small store? Usually not. Buffer-zone sizing has several parameters per SKU, and the benefit shows up in multi-stage chains with lumpy demand. Below a few hundred SKUs with one or two suppliers, a reorder point with a measured lead time gets most of the accuracy at a fraction of the maintenance.

Can Shopify do replenishment automatically? Not natively, as of November 2026. Shopify tracks quantities, purchase orders, and counts, but computes no per-SKU reorder points and places no orders on rules.

Next step

Write down which method each of your segments is actually using — not which one you intended. Then pick your ten most expensive SKUs, recompute velocity on in-stock days only, and see how many triggers move. That single correction improves whichever method you're running, because it fixes the input all four of them share.