OEM / private label · Multi-factory sourcing · Export project coordination
sales@moxasourcing.comCall UsChat on WhatsApp
Moxa Sourcing · Quantity · Cost · Inventory

Order quantities, inventory costs and price breaks

A larger order can reduce ordering frequency while increasing average stock. Compare the actual price-band rule and cost basis before using an inventory model to examine that trade-off. This guide explains ordering and holding costs, the assumptions of economic order quantity and the difference between all-units and incremental discounts; the model inputs remain separate from an actual supplier quotation.

How Moxa Sourcing helps

Specifications · Packaging · Quote scope

We coordinate product, packaging, sample and quotation-scope questions. Tell us the decision you need to resolve.

Line drawing of mugwort leaves

← MOQ, quotations and procurement cost

Three reading guides

01

MOQ and packaging: plan a defined trial order

Write down what is being ordered before asking for the minimum. The finished product, each variant, printed pouch, label, sleeve and retail box can have different quantity requirements and approval steps.

Which costs change when an order is split into smaller purchases?

Roberto Rossi’s Inventory Analytics distinguishes a fixed ordering cost from an inventory holding cost. The ordering cost is incurred each time an order is issued and is independent of that order’s size in the model. The holding cost is charged per unit of stock per time period. They respond differently to the purchasing pattern: splitting a requirement into more orders increases the number of ordering events, while holding stock for longer creates more exposure to the holding charge. [1]

Keep those model inputs distinct from a one-time product-development charge or a supplier’s minimum quantity. A fixed ordering cost repeats when another order is placed; a nonrecurring development cost need not do so. Likewise, a modelled holding cost requires a unit and a time basis, rather than an unexplained percentage. To use the comparison for a particular enquiry, identify the actual applicable charges and how each is incurred. The categories organise the analysis without supplying missing commercial terms. [1]

The model also distinguishes unit purchasing cost from ordering cost. Paying for the products and incurring the fixed cost of issuing an order are separate terms. A comparison that includes one but silently omits the other can favour a different purchasing pattern. Conversely, a charge already included in the agreed purchasing amount must not be added twice. The purpose of the breakdown is to identify which amounts depend on the number of orders, quantity purchased and time in stock. [1]

What does the basic economic order quantity model assume?

The basic economic order quantity model, or EOQ, assumes constant demand over time, continuous inventory review and immediate receipt of an order. All demand must be met when it occurs, so backorders are not allowed. The order quantity can vary continuously rather than being restricted to a supplier’s case multiples. These assumptions define the problem being solved. An observed supplier MOQ, a fixed carton size or uncertain demand is additional information that the basic formulation does not automatically incorporate. [1]

The notation separates demand rate from order quantity. Demand rate is an amount required per time period; order quantity is the amount received in one replenishment. They are not interchangeable. A monthly requirement and the quantity in each purchase may coincide, but that is one possible purchasing pattern rather than the definition of either input. Use one consistent time basis for demand and holding costs before comparing policies. [1]

Under immediate delivery, ordering at the point when stock reaches zero avoids unnecessary holding cost while meeting the model’s demand. This result depends on the stated assumptions. It is not a general instruction to wait until stock is exhausted before contacting a real supplier. A model with delivery lead time needs an earlier order point, and a model with uncertain requirements needs a treatment appropriate to that uncertainty. The basic EOQ result concerns the quantity trade-off within its defined setting. [1]

How do order quantity, stock coverage and order frequency relate?

With constant demand rate d and replenishment quantity Q, one replenishment covers Q/d time periods. Rossi calls the interval between successive orders the replenishment cycle. A larger quantity therefore gives a longer cycle in this model. This is a coverage calculation based on the assumed consumption rate. It does not establish how quickly a buyer will actually sell stock when demand is unknown or variable. [1]

The model’s stock falls evenly from Q to zero before the next immediate replenishment. Its average stock over that cycle is Q/2. That average differs from the amount initially received and from the amount remaining at a particular date. When comparing holding costs, use the stock measure that the calculation requires. Applying a charge to the full opening quantity throughout the entire cycle would represent a different stock pattern. [1]

Over a common time horizon, order frequency is demand divided by quantity per order. Increasing Q reduces the number of ordering events represented by that frequency, while increasing average stock. Rossi also defines an implied turnover measure as demand rate divided by average stock. The measures describe the same assumed replenishment pattern from different angles. Keep them distinct from observed turnover or a promised replenishment schedule in an actual order agreement. [1]

Which parts of total procurement cost matter to the EOQ comparison?

Rossi’s basic model combines three cost terms: fixed ordering cost per event multiplied by order frequency, unit purchasing cost multiplied by demand, and holding cost multiplied by average stock. Expressed over the same time basis, these give an average cost rate. Each term must have compatible units. An order’s product-purchase total is different from the average purchasing expenditure per time period, just as a fixed cost per order is different from the ordering-cost rate. [1]

When demand and unit purchasing price are constant and every required unit is bought, purchasing expenditure over a common horizon does not depend on the size of each order. The basic EOQ comparison can therefore remove that constant term while finding the quantity that balances ordering and holding costs. Rossi calls the remainder relevant cost. Removing a constant from this optimisation does not mean that purchasing expenditure disappears from a budget or a quotation. [1]

For positive fixed ordering cost, demand rate and holding cost, the basic optimum is the square root of twice ordering cost times demand divided by holding cost. Higher ordering cost favours larger replenishments; higher holding cost favours smaller ones; higher demand increases the optimal quantity under those assumptions. The result is a conditional model quantity. If unit prices change by quantity tier, or only certain quantities can actually be ordered, those conditions need to enter the comparison. [1]

How should a nearby feasible quantity be compared with the mathematical optimum?

Rossi provides a sensitivity relationship for quantities different from the basic EOQ. The ratio of relevant cost at another positive quantity to relevant cost at the optimum is half the sum of that quantity divided by the optimum and the reciprocal ratio. It compares ordering-plus-holding cost under the same basic assumptions. It does not compare every component of a supplier’s total quotation. [1]

The relationship makes the cost consequence of a quantity change explicit. Rather than assuming that any departure from the calculated optimum is unacceptable, evaluate the relevant cost at the feasible alternatives. The comparison still requires the same input basis and model scope. A case multiple, minimum order or practical replenishment interval may restrict the available choices, but the effect of such restrictions should be examined rather than described as an automatic commercial saving. [1]

Changing the quantity is different from changing the model’s inputs or conditions. The sensitivity relationship holds fixed the demand rate, ordering cost, holding cost and basic assumptions used to define the optimum. A different price tier, a different holding-cost basis or an uncertain demand pattern changes the problem. A result from the basic sensitivity expression should not be extended to those situations without revisiting the cost model. [1]

Does a quantity discount apply to every unit or only to the next band?

Compare the pricing rule before comparing unit prices [1]
Comparison pointAll-units discountIncremental discount
Where the lower rate appliesEvery unit in the qualifying orderOnly the quantity falling within that band
Purchasing amountWhole quantity multiplied by the applicable rateSum the quantity and rate for each used band
Meaning of the final band rateThe rate used for the whole qualifying orderThe marginal rate; average unit cost can differ

Rossi distinguishes all-units discounts from incremental discounts. With an all-units structure, reaching a quantity band makes that band’s unit purchasing price apply to the entire order quantity. With an incremental structure, the lower price applies only to the portion of the order lying in the relevant band. The same threshold and the same displayed lower rate can therefore describe different purchasing totals. The quotation needs to make the discount structure explicit. [1]

For an incremental discount, add the cost of the quantity in each completed band and the cost of the quantity in the final partially used band. Dividing that total by order quantity gives an average purchasing cost per unit. That average differs from the marginal rate charged for the last band. For an all-units discount, the applicable band rate applies to the whole quantity. Compare the resulting purchase totals before treating two tier schedules as equivalent. [1]

Quantity-dependent prices change the EOQ comparison because purchasing cost is no longer constant across all order quantities. Rossi’s discount model examines the cost within each price interval and compares the interval candidates. Its holding-cost input is expressed as a proportion of purchasing value, so the applicable price basis also affects holding cost. Keep that proportional input distinct from the absolute holding cost per unit per period used in the earlier basic model. [1]

A purchasing comparison needs both the pricing rule and the actual allowed quantities. Identify the sale unit, band thresholds, whether an endpoint qualifies for the next price and which charges remain separate. Those are quotation inputs to verify; the inventory model does not establish them for a supplier. A lower marginal or displayed unit price can be useful information, but it becomes a comparable total only after the applicable quantities and cost basis have been identified. [1]

Comparing waiting, shared events and period-specific inventory costs

What changes when a cost model permits demand to wait?

Rossi’s planned-backorder extension changes the requirement that every demand be met immediately. An unmet quantity remains due and is supplied when the next replenishment arrives. It is therefore different from a lost sale, where that demand leaves the system. A cost comparison must identify which situation it represents. Treating unfilled demand as if it simply disappeared would omit an obligation in a backorder model; treating a lost sale as eventual demand would describe a different problem. [2]

The extension chooses both a replenishment quantity and a maximum backlog. Immediately before replenishment, the net inventory level is negative by the outstanding backlog. The arriving quantity first clears that backlog; only the remainder becomes stock available for subsequent demand. Consequently, the quantity ordered is not the maximum positive stock held. Using the whole replenishment quantity as that peak would overstate the model’s holding-cost exposure. [2]

Planned backorders are a modelling assumption about eventual fulfilment, not an agreed customer tolerance. The mathematical treatment does not establish whether a particular buyer will accept delay, whether an order may legally be deferred, or what a supplier owes under an agreement. Those commercial conditions must be established separately before this kind of comparison can inform an actual purchasing decision. [2]

How are holding cost and waiting cost compared?

The penalty in this extension is charged for each unit backordered for each time period. It is not a single fee per late order or a percentage reduction in a quotation. A backlog of several units lasting several periods creates unit-time exposure. Changing from a daily to a monthly demand basis therefore also requires a compatible holding-cost and backlog-penalty basis. Otherwise, the apparent balance between keeping stock and making demand wait is an artefact of mismatched units. [2]

With constant demand and immediate replenishment, positive stock and outstanding backlog each form a triangular area over a cycle. Dividing those areas by the full cycle length gives their respective time averages. Rossi’s cost expression applies the holding rate to average positive stock and the penalty rate to average backlog, while retaining the ordering-cost term. A negative net inventory figure is not a negative physical holding cost that can cancel the charge for stock held earlier. [2]

For positive holding and penalty rates, the model’s optimal backlog as a share of replenishment quantity is the holding rate divided by the sum of holding and penalty rates. A higher penalty reduces that share. Substituting the share into the cost expression yields an effective holding coefficient equal to the product of the two rates divided by their sum. This is a mathematical reduction of the objective after optimisation; it does not lower an actual warehouse tariff. [2]

The optimal replenishment quantity in this extension is larger than the basic no-backorder EOQ for the same positive inputs. The extra quantity partly supports the planned backlog rather than an equal increase in peak physical stock. As the penalty becomes very large, the backlog share approaches zero and the quantity approaches the no-backorder result. These limiting relationships describe the model, without providing an observed cost of customer dissatisfaction. [2]

Which costs are shared when several items replenish together?

Rossi’s joint replenishment problem distinguishes a major setup cost for the system from a minor setup cost for each participating item type. The major cost is incurred whenever a replenishment event occurs, regardless of how many item types participate. An item’s minor cost is incurred only when that item participates. A comparison of combined and separate replenishment must therefore count events and participating items separately, rather than multiplying every charge by the number of product types. [2]

Combining two items at one event can remove a repeated major setup charge compared with two separate events, but it does not automatically remove either item’s minor charge. Nor does it make the items’ demand rates and holding costs identical. Each item can accumulate a different stock quantity and holding exposure. The meaningful comparison is the change in shared-event cost together with the changes in item-specific costs, on one common time basis. [2]

This distinction concerns cost incidence, not the physical design of a gift set or an instruction to combine shipments. A shared event is useful in the calculation only if the actual arrangement has the corresponding shared charge. Whether products can be ordered together, which services are included and which charges remain separate are commercial inputs. A mathematical grouping does not establish that a supplier offers those terms. [2]

Must every item use the same replenishment interval?

A common event structure does not require every item to replenish at every event. In the restricted joint replenishment policy described by Rossi, item cycles are selected from powers-of-two multiples of a base planning period. An item with a longer cycle can join only some of the more frequent events. The restriction coordinates timing while retaining different item frequencies; it is not a claim that every product should be purchased in the same quantity. [2]

For an item with a constant demand rate, a longer cycle increases the quantity replenished and the average stock associated with that cycle. Its setup cost occurs less frequently, while its holding-cost contribution increases. The shared major cost responds to the frequency of system events, so its incidence differs from a single item’s minor setup cost. Keeping these frequencies separate prevents an apparent saving created by charging a shared cost only once over an inconsistent horizon. [2]

The base period and the permitted cycle family constrain the comparison. A result found within that restricted family is a result for those allowed policies. It should not be described as an unrestricted optimum for every possible ordering pattern or as a guaranteed saving in a real procurement programme. Actual order calendars, allowed quantities and item availability still need to be represented when they affect feasible choices. [2]

How does changing demand affect the cost of buying ahead?

The dynamic lot-sizing setting replaces a constant demand rate with known requirements for separate periods over a finite horizon. Orders can be issued at period beginnings and arrive immediately in the basic formulation. The question for cost analysis is which periods one replenishment should cover. This is distinct from estimating uncertain future demand: variation between specified periods does not itself make the input a probability forecast. [2]

For a cycle beginning in one period and covering that period plus later periods, the fixed ordering charge is paid once. Units needed in the starting period incur no interperiod holding time in that cycle. Units for the next period are carried for one period; units for the following period are carried for two. Multiplying each later requirement by its holding duration and the per-unit, per-period rate gives the cycle’s holding cost. Adding the fixed charge gives the relevant cycle cost. [2]

This calculation explains why buying an entire horizon’s requirement at once can differ from buying each period’s requirement separately even when the unit purchase price is unchanged. The combined purchase avoids some ordering events but carries later requirements for longer. The comparison follows the actual sequence of period requirements, rather than substituting an average demand rate into a constant-demand formula and assuming the timing has been preserved. [2]

The basic formulation can omit constant unit purchasing expenditure when every required unit is purchased and the same unit price applies throughout the finite horizon. A time-dependent price, quantity discount or different set of fulfilled requirements changes that reasoning. Keep the relevant-cost calculation attached to its assumptions. Omitting a constant for comparison does not remove the purchase amount from the funds required to execute an order. [2]

How should a maximum order quantity affect the comparison?

Rossi’s capacitated extension imposes a maximum order quantity in each period while requiring demand to be met on time. A proposed low-cost replenishment that exceeds that maximum is infeasible. Replacing it with several permitted orders changes the number of fixed charges and may also change when inventory is held. The capacity limit belongs inside the cost comparison, rather than being checked only after an unconstrained purchasing pattern has been selected. [2]

A maximum quantity is different from a supplier MOQ: one caps what may be ordered in a period, while the other sets a minimum for an applicable order. Neither establishes the other. Where period requirements exceed what can arrive in that period, the formulation may require earlier stock, if earlier capacity and the starting stock permit it. An affordable-looking plan does not prove that all requirements can be met within those limits. [2]

Starting inventory also changes what remains to be purchased. Rossi discusses positive initial stock separately from the zero-stock basic case. Identify the opening quantity before treating every period requirement as a new purchase. The cost of carrying existing stock and the cost of placing additional orders are distinct contributions. This is an input distinction for a comparison; it is not evidence that a particular product is physically available or suitable for the intended order. [2]

FAQ

Which costs change when an order is split into smaller purchases?

Roberto Rossi’s Inventory Analytics distinguishes a fixed ordering cost from an inventory holding cost. The ordering cost is incurred each time an order is issued and is independent of that order’s size in the model. The holding cost is charged per unit of stock per time period. They respond differently to the purchasing pattern: splitting a requirement into more orders increases the number of ordering events, while holding stock for longer creates more exposure to the holding charge.

What does the basic economic order quantity model assume?

The basic economic order quantity model, or EOQ, assumes constant demand over time, continuous inventory review and immediate receipt of an order. All demand must be met when it occurs, so backorders are not allowed. The order quantity can vary continuously rather than being restricted to a supplier’s case multiples. These assumptions define the problem being solved. An observed supplier MOQ, a fixed carton size or uncertain demand is additional information that the basic formulation does not automatically incorporate.

How do order quantity, stock coverage and order frequency relate?

With constant demand rate d and replenishment quantity Q, one replenishment covers Q/d time periods. Rossi calls the interval between successive orders the replenishment cycle. A larger quantity therefore gives a longer cycle in this model. This is a coverage calculation based on the assumed consumption rate. It does not establish how quickly a buyer will actually sell stock when demand is unknown or variable.

Which parts of total procurement cost matter to the EOQ comparison?

Rossi’s basic model combines three cost terms: fixed ordering cost per event multiplied by order frequency, unit purchasing cost multiplied by demand, and holding cost multiplied by average stock. Expressed over the same time basis, these give an average cost rate. Each term must have compatible units. An order’s product-purchase total is different from the average purchasing expenditure per time period, just as a fixed cost per order is different from the ordering-cost rate.

How should a nearby feasible quantity be compared with the mathematical optimum?

Rossi provides a sensitivity relationship for quantities different from the basic EOQ. The ratio of relevant cost at another positive quantity to relevant cost at the optimum is half the sum of that quantity divided by the optimum and the reciprocal ratio. It compares ordering-plus-holding cost under the same basic assumptions. It does not compare every component of a supplier’s total quotation.

Does a quantity discount apply to every unit or only to the next band?

Rossi distinguishes all-units discounts from incremental discounts. With an all-units structure, reaching a quantity band makes that band’s unit purchasing price apply to the entire order quantity. With an incremental structure, the lower price applies only to the portion of the order lying in the relevant band. The same threshold and the same displayed lower rate can therefore describe different purchasing totals. The quotation needs to make the discount structure explicit.

Does a backorder model establish permission to deliver late?

No. A planned-backorder model assumes eventual fulfilment and assigns a cost to units waiting over time. Actual customer acceptance, contractual obligations and applicable conditions must be established separately.

Comparing a minimum or quotation?

Ask about the product, packaging minimum, quoted unit or delivery scope. Start with your question; the current page is included.

Sources and scope

Adapted and restructured from Roberto Rossi, Inventory Analytics (Cambridge, UK: Open Book Publishers, 2021), Deterministic Inventory Control, printed pp.39–46 and 50–52. Copyright © 2021 Roberto Rossi. Source licensed under Creative Commons Attribution 4.0 International. Moxa Sourcing edited these buyer-focused explanations and prepared the Arabic translation; source figures, code and numerical examples are not reproduced. The source is supplied without warranties, and its author or publisher does not endorse this adaptation. Reviewed 20 November 2026.

  1. Roberto Rossi, Inventory Analytics (2021) · DOI: 10.11647/OBP.0252 · CC BY 4.0
  2. Roberto Rossi, Inventory Analytics (2021), inventory-cost extensions · DOI: 10.11647/OBP.0252 · CC BY 4.0

    Roberto Rossi, Inventory Analytics (Cambridge, UK: Open Book Publishers, 2021), Deterministic Inventory Control, selected explanations in printed pp.53–54, 59–64 and 66–68. Copyright © 2021 Roberto Rossi; licensed under Creative Commons Attribution 4.0 International. Moxa Sourcing adapted the inventory-cost explanations and prepared the Arabic translation. Source code, figures and numerical examples are not reproduced; the comparison does not establish actual demand, charges or permission to delay orders. The source is supplied without warranties; its author and publisher do not endorse this adaptation. Reviewed 20 November 2026.

Chat on WhatsApp+852 5725 4301