The last few units needed to reach a discount threshold can change the price of an entire order. That makes a volume offer different from an ordinary comparison between two unit prices. Sometimes buying more than the immediate requirement reduces the total invoice. Sometimes the apparent saving disappears once the cost of the excess is included. The answer depends on the pricing rule, not simply on the percentage displayed beside the offer.
On 26 June 2026, Vedomosti reported that Russian Railways in Russia had introduced discounts for specified industrial freight movements, with the stated aim of attracting additional cargo. That announcement is a reminder to examine what a discount applies to; it does not establish the threshold arrangements used below.
This article compares two independent purchasing examples. Neither is a reconstruction of a railway tariff or a claim about the company's contracts. All quantities, prices and costs are invented. The examples isolate a narrow decision: whether a buyer with a known requirement should order exactly that quantity or increase the order to a stated price boundary.
Write down the complete offer
Imagine a buyer that needs 95 identical units for one defined period. Each useful unit serves the same purpose, and buying fewer than 95 is not an acceptable option. Quantities must be whole numbers. There is no uncertainty about the requirement in the initial calculation, and all goods can arrive at the same time. Differences in quality, delivery and supplier reliability are deliberately excluded.
The first offer charges ten monetary units for each item when the order contains fewer than 100. At 100 or more, the price becomes nine for every item in the order. This is an all-units discount: reaching the threshold changes the price applied to the complete qualifying quantity, including units that would have cost ten in a smaller order.
The second offer will use the same lower price of nine, but only for the hundredth item and later items. The first 99 remain priced at ten. Stating that convention explicitly matters. An expression such as discount above 100 could mean something different, so the arithmetic cannot begin until the first discounted unit and the treatment of earlier units are unambiguous.
More units can produce a smaller invoice
Under the all-units offer, ordering the required 95 costs 950. Ordering 99 costs 990. Ordering 100 costs only 900, because the lower price applies to all 100. Ordering 101 costs 909. The invoice rises within each price band but drops when the buyer moves from the last quantity below the threshold to the first quantity that qualifies.
The move from 99 to 100 therefore changes the invoice by minus 90. It is not correct to describe that particular step as an additional payment of nine. The new item costs nine, but the previous 99 items each receive a one-unit reduction. Those 99 units of savings more than offset the nine charged for the added item.
This does not mean the supplier pays the customer to take goods. Both invoices remain positive, and they describe alternative orders, not a refund transaction between completed purchases. The negative difference is a comparison across the stated pricing boundary. It arises because the complete invoice is recalculated under another price band.
Compare the threshold with the actual requirement
The buyer's decision starts at 95, not 99. Increasing the order to 100 adds five items while reducing the purchase bill from 950 to 900. Before any separate consequences of the extra goods, the larger order saves 50. A decision based only on avoiding unnecessary quantity would miss that feature of the offer.
Conversely, a decision based only on the nine-unit price would be incomplete. The buyer must purchase 100 units to obtain it. Multiplying the required 95 by nine would produce 855, but that is not an available invoice under the stipulated offer. A conditional price cannot be detached from the quantity needed to qualify for it.
The excess needs its own economic treatment
In this first example, only 95 items are useful within the chosen period. The additional five have no resale value and cannot replace a later purchase within that same horizon. This is a deliberate assumption, not a general claim that excess stock is worthless. It lets the comparison focus on the saving and the burden associated with obtaining it.
Suppose each excess item creates a separate cost of six for handling and eventual disposal. These six units are additional to its purchase price, not another allocation of the same invoice. Ordering 100 now costs 900 plus five times six, or 930. Ordering 95 still costs 950. The larger order remains cheaper, but its advantage has fallen from 50 to 20.
If the extra burden is twelve per excess item, the larger order costs 960. Ordering exactly 95 is now cheaper by ten. The purchasing rule has not changed. What changed is the consequence of keeping items that do not satisfy an additional useful requirement. The decision must include that consequence rather than treat the invoice as the complete economic result.
Calculate the break-even burden
The threshold order saves 50 on purchase cost and creates five excess items. It can therefore absorb ten of additional cost per excess item before the alternatives become equal. At exactly ten, the larger order costs 900 plus 50, or 950. The comparison provides no cost preference between the two quantities under those assumptions.
Below ten, ordering 100 is cheaper; above ten, ordering 95 is cheaper. This is a threshold within the model, not a universal allowance for storage or waste. It follows from the specified need, the offered prices and the quantity boundary. A different requirement or discount changes both the saving and the number of excess items over which it must be spread.
The calculation also requires consistent cost boundaries. If a quoted disposal charge already includes handling, adding another handling amount for the same activity would count it twice. If a genuine additional cost is omitted, the apparent saving is overstated. The purpose is to compare complete alternatives, not to assemble as many cost labels as possible.
An incremental discount behaves differently
Now apply the second offer. The first 99 units cost ten each, and only unit 100 onward costs nine. Ordering 95 still costs 950. Ordering 100 costs 990 plus nine, or 999. Ordering 101 costs 1,008. There is no downward jump in the total bill because none of the earlier units becomes cheaper when the threshold is crossed.
Moving from 99 to 100 adds nine to the invoice. Moving from 95 to 100 adds 49: four units at ten and one at nine. With no useful value for the excess and no negative additional costs, buying more cannot improve the buyer's result in this version. The minimum adequate quantity remains the cheapest available choice.
Both offers can advertise a reduction from ten to nine at a volume boundary. That shared percentage does not make them commercially equivalent. One changes the price of the entire order; the other changes only the price of later units. A buyer needs the full calculation rule, not an inference drawn from the most prominent number.
Only a few quantities need comparison here
Under the all-units offer, quantities from 95 through 99 have the same unit price of ten. With a non-negative additional burden for excess goods, each extra unit in that band raises total cost. The cheapest adequate quantity within the lower band is therefore 95. There is no reason within the example to choose 96, 97, 98 or 99.
At 100, the price changes. Beyond 100, every extra unit adds nine of purchase cost plus the specified excess burden. Total cost rises again, so 100 is the cheapest quantity in the upper band. Comparing 95 with 100 is sufficient to identify the minimum across the allowed quantities in this particular model.
This is not a claim that every quantity-discount problem reduces to two candidates. Additional price bands would introduce additional boundaries. Minimum packaging sizes might remove some quantities from consideration. A positive useful value for later units could change the objective. The short comparison works because the current assumptions make cost increase within each of two clearly defined bands.
Distance from the boundary matters
Change the immediate requirement to 80 while leaving the all-units offer unchanged. Buying exactly what is needed costs 800. Buying 100 costs 900 even before any extra handling or disposal. The lower unit price no longer compensates for the larger quantity. The existence of a discount therefore does not create a general rule to purchase up to its threshold.
The earlier result depended on being sufficiently close to the boundary. At 95, the threshold order had a smaller invoice. At 80, it has a larger one. A purchasing team cannot transfer the first conclusion to every order that mentions the same discount schedule. The actual requirement determines which alternatives need comparison and what their excess quantities are.
It is also important not to define the requirement backwards from the offer. Calling the extra items necessary merely because they unlock a price reduction hides the trade-off. The useful requirement should be established separately; the calculation can then reveal whether deliberately buying beyond it is worthwhile under the stated costs.
Keep cash paid separate from goods consumed
The comparison concerns the total cost of alternative orders within one specified horizon. It does not establish how a real business should recognise inventory or expenses in its accounts. Buying 100 items and using 95 are different physical events. A model that assumes no value for the remainder must disclose that assumption instead of presenting it as an accounting rule.
If the extra five will replace a known future purchase, the comparison needs another time boundary and the relevant future alternative. If they may be sold, that possibility needs a supported receipt and any selling cost. Those changes might strengthen or weaken the case for a larger order. They cannot be added as benefits without also checking the assumptions that make them available.
Affordability remains separate as well. The arithmetic ranks the stated alternatives but does not prove that the buyer can pay, store or accept either delivery. Here those operational conditions are assumed feasible. A real decision must establish them before treating the lowest calculated cost as an executable purchasing plan.
Read the qualifying quantity as carefully as the price
The example defines one order as the qualifying unit. A different agreement might refer to an accumulated quantity over a period or to another explicitly defined scope. That would be a different calculation. The buyer should not combine separate orders, products or delivery dates merely because doing so would cross the numerical threshold on a spreadsheet.
Returns could also matter if the applicable terms change eligibility when quantities are reduced. This article does not presume any such rule or advise using returns to obtain a discount. It identifies the boundary that would need checking. The relevant question is which quantity the actual agreement recognises, not which total makes the proposed saving look largest.
Where the conditions are unclear, calculating several interpretations can expose the ambiguity but cannot settle it. The parties need an agreed description of the offer. A precise numerical answer to an unspecified contract is still unsupported. Clear wording about earlier units, later units and the qualifying order prevents a small difference in interpretation from becoming a large difference in expected cost.
A compact comparison makes the decision visible
A useful record can preserve the essential inputs without disguising the choice behind an average price:
- State the independently established useful requirement.
- Identify the qualifying quantity and first discounted unit.
- Specify whether the lower price applies to all units or only later units.
- Calculate the complete invoice for each feasible candidate order.
- Record the value and separate costs of any excess quantity.
- Compare total outcomes within one consistent horizon.
The final result should show both the invoice and the excess burden. In the 95-unit example, the all-units threshold lowers purchase cost by 50. A burden of six per excess item leaves a saving of 20; a burden of twelve makes the larger order worse by ten. Under the incremental offer, the larger order never has the initial invoice saving at all.
Ordering near a discount boundary is therefore a question about the shape of the complete payment schedule. The buyer must know which previous units are repriced and what happens to the additional goods. Once those details are explicit, neither a striking percentage nor a general dislike of surplus stock needs to decide the answer. The available orders can be compared directly, on the cost of meeting the actual requirement.