Tiered Pricing Calculator

How it is calculated

Tiered pricing charges different rates for different quantity bands, and there are two structures that produce very different bills. Marginal tiering charges each band its own rate, the first 100 units at one price, the next 400 at another. Retroactive tiering applies the rate of the highest band reached to the entire quantity.

The distinction is not a detail. On the same schedule, retroactive tiering can make a customer's bill fall as they buy more, which creates a cliff at every threshold.

Numbers on it

A schedule of $10 up to 100 units, $9 from 101 to 500, $8 above 500. A customer buying 501 units.

Marginal: 100 × $10 + 400 × $9 + 1 × $8 = $4,608. Retroactive: 501 × $8 = $4,008.

Now look at the cliff. Under retroactive tiering, a customer buying 500 units pays $4,500 and one buying 501 pays $4,008, $492 less for one more unit. Every customer near the boundary discovers this, and the ones who do not feel cheated when they find out.

What it does not tell you

Retroactive tiering is more motivating and considerably more expensive, and the cost is concentrated exactly where volume is largest. Modelling it against your actual order distribution rather than a single example is essential, because the cost depends entirely on how many customers sit just above each threshold.

Marginal tiering has no cliff and correspondingly less pull. A customer at 480 units has little reason to reach 501, because only the marginal unit gets the better rate.

What follows from it

Model both against last year's order distribution before choosing. The answer depends on your customers, not on which structure sounds better in principle.

If you use retroactive tiering, place the thresholds where few customers already sit. A threshold just above a cluster of existing orders gives away the discount to volume you already had, which is the worst of both structures.

Tiering as a segmentation tool

The commercial purpose of tiering is not to reward volume but to charge different customers different prices without appearing to. A small buyer and a large one pay different rates, both feel fairly treated, and neither has grounds for complaint because the schedule is public.

That framing changes how the tiers should be designed. The question is not what discount volume deserves but where the customer segments actually divide, and those boundaries are usually visible in the order data as clusters rather than as a smooth distribution.

Setting thresholds at the gaps between clusters captures the segmentation cleanly. Setting them at round numbers, which is what most schedules do, cuts through the middle of a segment and gives the discount to roughly half of it for nothing.

Publish the schedule. A tiered structure that customers can see lets them decide to reach the next tier, which is the entire mechanism. A schedule available only on request is a negotiation position rather than a pricing structure, and it produces neither the segmentation nor the volume pull that tiering exists to deliver.

Published schedules also remove an enormous amount of sales admin. Every question about whether a particular quantity qualifies for a particular rate is answered by the table rather than by someone checking, and the consistency that produces is worth more than the flexibility it gives up.