Inventory Replenishment Methods Compared

Most brands do not choose a replenishment method. They inherit one, usually whatever the first operations hire set up, and then apply it to every product regardless of how differently those products behave. That is the actual problem, not which method is best, because no single method is right for a whole catalog.

The four common replenishment methods are the reorder point, periodic review, min/max, and vendor managed inventory. They differ on when you check stock and what triggers an order. Match each method to product value and demand volatility rather than standardising on one, and most catalogs end up running two or three.

Key takeaways

• The real choice is continuous versus periodic: everything else follows from whether you watch constantly or check on a schedule.
• Method should track product value: precision costs attention, and attention is your scarcest input.
• Running one method catalog-wide is the common error: it over-serves the tail and under-serves the products that matter.
• Volatility decides as much as value: a cheap but erratic product can justify more precision than an expensive steady one.

What are the four replenishment methods?

Each answers the same two questions, when do we look and what makes us order, with a different trade between precision and effort.

• Reorder point. when you check: Continuously; what triggers the order: Stock hits a calculated trigger; order quantity: Set separately; best fit: Higher-value or volatile products
• Periodic review. when you check: On a fixed schedule; what triggers the order: The review date arrives; order quantity: Top up to a target; best fit: Consolidated ordering with one supplier
• Min/max. when you check: Continuously; what triggers the order: Stock hits the min; order quantity: Refill to the max; best fit: Steady, low-value products
• Vendor managed. when you check: Supplier monitors; what triggers the order: Supplier decides inside agreed bands; order quantity: Supplier decides; best fit: Deep supplier relationships, high volume

The first three are the ones most eCommerce brands will actually use. Vendor managed inventory is included because it comes up in every comparison, but it needs a supplier relationship and order volume most growing brands do not yet have, and suppliers rarely offer real programmes at modest volumes.

Continuous or periodic: the choice underneath the others

Strip the labels away and there are two philosophies. Continuous review watches stock all the time and orders the moment a threshold is crossed, so the timing follows the product's actual consumption. Periodic review checks on a calendar and orders whatever is needed at that moment, so the timing follows your schedule.

Continuous is more responsive and more precise, and it costs more attention, or a system that can do the watching. Its failure mode is that if the trigger is stale, it fires confidently at the wrong level and nobody notices, because the mechanism is working exactly as configured.

Periodic is simpler, easier to staff, and it consolidates ordering neatly, which matters when freight or supplier minimums reward combining products into one purchase order. Its failure mode is structural: a product can cross its trigger the day after a review and sit exposed until the next one. That gap is the price of the simplicity, and it is why the review interval has to be set against your lead times rather than for convenience.

Rule of thumb: if a product could sell through its buffer within one review interval, it does not belong on periodic review. Move it to continuous or shorten the interval for that group.

How do you match a method to a product?

Two dimensions decide it: how much the product is worth and how much its demand moves. Value determines how much precision is worth paying for. Volatility determines how much precision is actually needed.

High value, high volatility

Calculated reorder points, reviewed frequently. These are the products where a stockout costs real revenue and where demand moves enough that a fixed level goes stale quickly. They justify the maintenance: recalculated buffers, verified lead times, and a check that runs more often than monthly. This group is usually small, which is what makes the effort affordable.

Steady products, whatever their value

Min/max, and leave it alone. A product selling roughly the same amount every week from a reliable supplier does not need a recalculated trigger, because the number that was right last quarter is still approximately right. The saving here is not on the ordering, it is on the attention, and that attention is what funds the precision applied to the first group.

Low value, any pattern

The simplest rule that works, reviewed rarely. The temptation is to hand-manage these when one of them stocks out, and resisting it is the whole discipline. You accepted a slightly worse service level on low-value products so that your best sellers could have a better one. Reversing that one product at a time undoes the trade. Sorting products into these tiers is what ABC analysis is for, and the min/max mechanics are covered in min/max and par levels.

What actually goes wrong with each method?

Knowing the failure mode is more useful than knowing the definition, because every method works on paper.

Reorder points fail quietly through staleness. The trigger was correct when it was set, demand grew 30%, the supplier drifted from nine days to twelve, and now it fires three days too late on every cycle with no error message. Periodic review fails through the gap between checks, which is invisible until something sells through the buffer in the middle of a cycle. Min/max fails through ceilings nobody revisits, which quietly become overstock machines because the refill happens regardless of whether demand still justifies that level. Vendor managed fails through loss of visibility, since you have handed the decision to someone whose incentive is to ship you more rather than less.

Three of those four failures share a cause: a number set once and never revisited. That is the argument for automating the maintenance rather than the ordering.

Conative AI's Buying Agent works against current numbers rather than inherited ones. It analyses out-of-stock and overstock risk per product, estimates the revenue at stake, and drafts the purchase order matched to your lead times, minimum order quantities, and supplier terms, so the trigger reflects this month's demand rather than last quarter's. It drafts and recommends; it does not place the order or change your product mix on its own. Book a call to see it run against your own catalog on the inventory planning platform.

Frequently asked questions

Which replenishment method is best for eCommerce?

None of them universally, and standardising on one is the most common mistake. Most eCommerce catalogs end up running reorder points on the products that carry revenue and min/max on the long tail, with periodic review where consolidating orders with a supplier pays. The mix is the answer.

What's the difference between continuous and periodic review?

Continuous review keeps a constant watch and fires as soon as a product drops past its threshold. Periodic review looks only on set dates and tops up whatever it finds low at that moment. Continuous is more responsive and needs more attention or automation; periodic is simpler but leaves a gap between checks.

Can you use more than one replenishment method?

Yes, and you probably should. Products differ enough in value and volatility that one method will over-serve some and under-serve others. Running two or three across a catalog is normal practice, and the split usually follows an ABC-style classification.

Is vendor managed inventory worth it for a small brand?

Rarely, at least at first. It needs order volume and a supplier relationship that most growing brands have not yet built, and suppliers seldom offer genuine programmes at modest volumes. Automating your own replenishment usually delivers a similar result while keeping visibility and control in-house.

How do you know when to change replenishment method?

When the current method's failure mode starts showing up regularly: repeated late triggers suggest a stale reorder point, stockouts landing mid-cycle suggest the review interval is too long, and creeping overstock on steady products suggests a min/max ceiling nobody has revisited. The symptom points at the method.

Does the method change the order quantity?

Sometimes. Min/max sets the quantity implicitly as the gap between current stock and the ceiling, and periodic review tops up to a target. Reorder points say nothing about quantity, so it comes from a separate rule such as economic order quantity or a supplier minimum.