SaaS Pricing Models, Explained

Per-seat pricing is the most forecastable model and the easiest to sell, because buyers can compute their bill exactly. It aligns revenue with headcount, which is a proxy for value in collaboration tools and an increasingly poor one elsewhere.

Its structural weakness is that customer success does not expand revenue unless the customer also hires. It also creates a direct incentive for buyers to limit access, which suppresses the adoption that drives retention.

Usage pricing aligns cost with value delivered and removes adoption friction: customers pay nothing until they use something. Infrastructure and AI companies adopted it because their own costs are consumption-driven.

The cost is predictability on both sides. Customers cannot budget and vendors cannot forecast, which is why almost every mature usage-based company sells annual commitments with overage rather than pure pay-as-you-go.

The dominant enterprise pattern is a committed annual contract with usage drawn down against it and overage billed above. Vendors get forecastable revenue, customers get volume discounts and a budget line.

The negotiation shifts to commitment sizing, where the vendor's interest is a large commitment and the customer's is an accurate one. Unused commitment is pure margin for the vendor and pure waste for the buyer.

Tiering by capability rather than volume works when the gated features map to genuine segment differences: audit logs and SSO for enterprises, basic functionality for individuals.

It fails when critical security features sit behind the most expensive tier, a practice that has drawn sustained criticism from buyers and now factors into procurement decisions at security-conscious organisations.