skip to content

On a graduated volume schedule, what does crossing into a cheaper tier reprice, and what did you promise to get it?

level: middleimportance: nice to knowfreq 30%

answer

  1. a discount with no promise
  2. the schedule, not a contract
  3. graduated against whole-period
  4. marginal falls, average lags behind
  5. only units above the threshold

basics

~20 s

On a graduated schedule only the units above the threshold get the cheaper rate; earlier units keep the rate of the tier they fell in. Nothing is promised — the rate falls automatically with consumption inside a billing period.

solid answer

~40 s

A tiered volume rate is the one price reduction on a platform that costs no promise at all. The published schedule says the first block of units in a billing period is charged at one rate, the next block at a lower one, and so on, and it applies by itself as consumption accumulates. On a **graduated** schedule the cheaper rate is marginal: it applies only to the units above the threshold, while earlier units keep the rate of the tier they landed in. That makes the **average** rate a weighted blend that always lags the current **marginal** rate, which is the number teams most often quote by mistake when forecasting. The schedule also resets with each billing period, so a month that started slowly starts again at the top of the schedule.

code

pseudocode · 22 lines
pseudocode
monthlyUnits = 900

tiers = [
  { upTo: 100,       ratePerUnit: 1.00 },
  { upTo: 500,       ratePerUnit: 0.80 },
  { upTo: unbounded, ratePerUnit: 0.60 }
]

charge      = 0
remaining   = monthlyUnits
floorOfTier = 0

for each tier in tiers:
    unitsInTier = min(remaining, tier.upTo - floorOfTier)
    charge      = charge + unitsInTier * tier.ratePerUnit
    remaining   = remaining - unitsInTier
    floorOfTier = tier.upTo
    if remaining == 0:
        stop

// 100*1.00 + 400*0.80 + 400*0.60 = 660
// average rate 0.733 per unit, next unit costs 0.60

go deeper

for a junior

Know that some resources simply get cheaper per unit as you consume more in a billing period, with nothing signed and no term involved.

for a middle

Explain the graduated mechanism: the cheaper rate applies to units above the threshold, so the average rate paid always trails the current marginal rate.

for a senior

Show that you forecast on the blended average rather than the bottom tier, and that you read whether a given schedule is graduated or reprices the whole period.

for a principal

Judge when consolidating consumption to reach a tier is worth the coupling it creates, and when a commitment already covers so much usage that the tier never arrives.

## A discount with no promise attached Most ways of paying less on a platform require you to give something up: a term, a prepayment, a fixed resource shape, or the right to keep a machine you were lent. A **tiered volume rate** is the exception. It is written into the published schedule for a resource, it applies automatically as consumption accumulates inside a billing period, and nothing is signed, reserved or promised. That also means it protects nothing. It is a property of the price list, not an agreement, so it can be changed by the provider and it gives you no claim if your consumption falls. ## How a graduated schedule computes a charge Take an invented schedule for a metered unit and a month that ends at 900 units: | Tier | Units in tier | Rate per unit | Charge | |---|---|---|---| | Up to 100 | 100 | 1.00 | 100 | | 100 to 500 | 400 | 0.80 | 320 | | Above 500 | 400 | 0.60 | 240 | | **Total** | **900** | | **660** | The month costs 660 for 900 units. Two different rates can be read off that, and they answer different questions: - The **marginal rate** is 0.60 — what the next unit costs right now. It is the number to use when asking whether an extra workload is worth running this month. - The **average rate** is 660 / 900, about 0.733 — what the month actually cost per unit. It is the number to use when forecasting a similar month. Quoting the marginal rate as if it were the average is the classic error, and it understates a forecast by the whole weight of the upper tiers. ## Graduated against whole-period schedules Not every volume schedule is graduated. Some are written so that reaching a threshold reprices **the entire period's consumption** at the lower rate, which produces a cliff rather than a slope: one extra unit can lower the whole bill. Providers differ here, and so do different resources on the same platform, so the schedule has to be read rather than assumed. The distinction matters in two places: - **Forecasting.** A graduated schedule blends; a whole-period one steps. - **Behaviour near a threshold.** Under a whole-period schedule there is a genuine incentive to push consumption over the line before the period closes. Under a graduated one there is none, because the units below the threshold never reprice. ## What a volume tier is not - **It is not a commitment.** No term, no promised quantity, no shortfall to pay. Consumption that does not happen simply is not charged. - **It does not stack as a second discount on committed usage in the obvious way.** Usage already covered by a commitment is priced by the commitment; it is the uncovered remainder that walks the volume schedule. A large commitment can therefore keep the remaining metered usage below the threshold that would have earned a cheaper tier. - **It is not a negotiated rate.** It is published and applies to everyone consuming at that level. - **It does not carry across periods.** The schedule resets when the billing period does. ## Reading it in an interview The askable point is the direction of the repricing and the marginal-against-average distinction, because both are easy to state backwards and both change a forecast materially. If you are asked what a volume tier costs you, the honest answer is nothing directly — the cost is in the planning error it invites, when a team builds a business case on the bottom-tier rate and then meets the top of the schedule again on the first day of the next period.

  • Why can a volume tier make a cost forecast look better than it turns out?
    Because the tier is reached by cumulative consumption inside one billing period, and the schedule resets with the period. A forecast built on the cheapest marginal rate assumes every unit next period is priced at the bottom of the schedule, when the early units are charged at the top of it again.
  • How does a volume schedule interact with a commitment you already bought?
    They price different usage. Usage matched by the commitment is billed at the committed rate, and only the uncovered remainder walks the volume schedule. A large commitment can therefore hold metered consumption below the threshold that would have earned the cheaper tier.

saying these in an interview costs you the question

  • Thinks a volume tier requires signing a term commitment first
  • Assumes every volume schedule reprices all units once a tier is crossed
  • Quotes the marginal tier rate as the average rate paid
  • Confuses a published volume tier with a negotiated account discount
  • Forgets the schedule resets with each billing period