Tiered data licenses: test what happens at the quantity boundary

Name the licensed unit and the tier method

A quoted tier table can hide a material receipt difference. Before comparing offers, specify what one unit means, the measured period and whether the reached rate applies to every unit or only the units in that band. A price per record is especially unhelpful if the agreement actually licenses seats, projects or deliveries.

Stripe’s public tiered-pricing documentation distinguishes volume pricing, which applies the reached tier’s rate to the whole quantity, from graduated pricing, which charges each band separately. Its pricing-model documentation also distinguishes recurring and usage-based arrangements. These are vendor arithmetic descriptions checked 11 October 2026, not observed data-license prices or a recommendation to adopt a billing platform.

Calculate both methods at the boundary

This hypothetical nonexclusive license covers one quarter of use by named internal project seats. Its invented schedule is USD 1,000 per seat-quarter for the first ten seats and USD 700 beyond ten. Under the volume alternative, reaching eleven seats applies USD 700 to all eleven. Under the graduated alternative, only the eleventh and later seats use that rate. Both examples assume the same permitted use and no minimum.

Ten seats cost USD 10,000 either way. Eleven cost USD 7,700 under volume pricing and USD 10,700 under graduated pricing. Adding one seat therefore reduces the volume total by USD 2,300 but increases the graduated total by USD 700. At twenty-five seats, the totals are USD 17,500 and USD 20,500. The descending volume boundary is an effect of the stated formula, not an error that a reviewer can silently repair.

Make the quantity independently reviewable

The agreement needs an operational definition of a seat-quarter: for example, the authorized seats on one agreed quarterly roster rather than a sum of daily logins. Define additions, substitutions, affiliate access, suspended accounts and the cutoff. This article’s invented roster contains eleven seats; it does not establish what a real agreement should count or prove any user activity.

Retain the approved roster version, reporting period, report review and correction process. A sum of event logs can count the same person repeatedly, while a peak concurrent count answers a different question. Choose the actual contract measure before applying a rate. State whether a zero-quantity period owes nothing, a minimum or a fixed fee, and keep taxes and currency conversion separate from the quantity formula.

Separate receipts from preparation and conditional costs

For the eleven-seat hypothetical, assume USD 2,000 of preparation is incurred before a deal; USD 1,500 of delivery engineering occurs only after the license; and a USD 100 external payment-processing charge applies only if a receipt arrives. Assume the agreed quarterly report is accepted and both alternatives pay their calculated amounts. The remaining cash after those stated costs is USD 4,100 for volume or USD 7,100 for graduated pricing, before tax and omitted costs.

If no license occurs, no receipt, delivery engineering or payment-processing charge occurs, and the preparation outflow is USD 2,000. This is one possible paid license, not a pipeline or an expected-value forecast. The invented processing charge is not a VOID seller referral fee: VOID charges no upfront seller referral fee and may receive conditional compensation from a receiving program. Disclose an applicable arrangement separately; do not add an unagreed deduction to the license.

OpenStax’s managerial-accounting text distinguishes costs that change between alternatives from already-incurred costs. Preparation stays visible in this whole-project cash view; it is not a new outflow caused by choosing one paid tier instead of the other.

Put a reproducible schedule beside the offer

The owner’s comparison brief should include the unit definition, tier method, boundary examples, zero/minimum treatment, reporting evidence and payment assumptions. A statement such as discounted above ten is incomplete until the all-units or marginal treatment is explicit. Ask the counterparty to confirm a completed eleven-seat example before treating either schedule as an offer you can compare.

Offer comparison can retain scope and payment differences; Earnings can model the receipt and costs as an explicitly hypothetical scenario. Those tools do not reproduce every tier convention, so attach the calculation table. Permission for a named-recipient introduction, sample evaluation and the final license remain separate. No amount here is market evidence of what an archive is worth or what its owner can expect to earn.