Choose the pricing unit before the price

Founders often ask whether the price should be $49 or $99 before deciding what the customer is buying. That reverses the work. A pricing unit determines what expands the bill, how predictable spend feels, which behavior gets discouraged, and whether revenue grows alongside delivery cost.

The economics chapter's contribution-margin treatment is useful context at metrics and economics chapter. This guide does not recommend a current market price.

Evaluate the unit, not the packaging

A unit should pass five tests:

• Value alignment: more units usually mean more customer value.
• Cost alignment: supplier cost does not grow much faster than revenue.
• Measurability: both sides can audit the count.
• Forecastability: the buyer can budget without fear.
• Controllability: the buyer can influence consumption and avoid surprise.

Stripe's billing documentation describes flat-rate, per-seat, tiered, and usage-based models; it is vendor documentation about structures the product supports, not evidence that one structure performs better (Stripe pricing models).

Compare three common choices

Seat pricing works when each enabled person receives recurring value and headcount is visible. It becomes awkward when a small number of operators create value for a large organization, or when customers avoid inviting collaborators to contain cost.

Usage pricing works when consumption is measurable and is a fair proxy for value or cost. It transfers volume risk to the buyer and can make adoption feel dangerous. Google Cloud Marketplace documentation requires usage products to define measurable metrics and reporting units (Google Cloud pricing models). That is marketplace implementation guidance, but it exposes a broader truth: a vague 'unit' is not billable trust.

Outcome pricing links payment to a result such as a recovered payment or qualified transaction. It can align incentives, but only when attribution, baseline, timing, exclusions, and customer contribution are contractible. Many important outcomes are jointly produced; arguing about causation can consume the margin.

Hybrid models can combine a base commitment with usage. They reduce supplier downside but add explanation and billing complexity.

Worked hypothetical comparison

Hypothetical: A document-review product processes supplier contracts for procurement teams.

• Per seat: $180 per reviewer per month. Easy to budget for six reviewers, but discourages occasional legal and finance collaborators.
• Per document: $14 per completed review. Tracks processing cost, but a 40-page bespoke agreement and a two-page renewal consume different effort.
• Per accepted exception found: $120. Sounds value-linked, but acceptance depends on customer policy and reviewer judgment.

The team chooses a base subscription that includes a document allowance plus a clearly defined overage. It then tests whether 'completed review' needs a complexity band. The figures are fictional; the decision comes from unit behavior, not competitive price matching.

Run a shadow-billing test

For four weeks, calculate invoices under two or three candidate units without charging them. For each account, record bill variance, gross contribution, customer value proxy, disputes, metering failures, and behavior that the unit would encourage. Share sample bills with prospects and ask them to forecast next month's amount. If they cannot, the unit may be commercially costly even when mathematically elegant.

Pricing-unit decision sheet

• Unit definition and exclusion rules
• Customer action that creates one unit
• Value hypothesis and cost driver
• Meter owner, audit trail, and dispute process
• Expected monthly range for low, normal, and high use
• Gross contribution under each range
• Behavior the unit rewards or suppresses
• Procurement and billing-system fit
• Conditions for revisiting the unit

Limits

A good unit does not determine willingness to pay, discount policy, contract length, tax treatment, or revenue recognition. Outcome pricing can create legal and measurement questions, and usage pricing may require controls for surprise bills. Test with buyers and qualified advisers where contracts, regulated outcomes, or accounting judgments are material.