Buttondown, the email newsletter tooling vendor, has published an analysis of usage-based pricing versus seat-based pricing, and its framing matters for anyone running a small product: the question is not which model is universally better, but which failure mode you can tolerate at your current scale.

The piece — titled "Usage-based vs. seat pricing: what the data says for small products" — approaches the question from the vendor's own position as a small product competing against larger players. That vantage point shapes everything that follows, and product managers at early-stage companies should read it with that context in mind.

The two models, precisely defined. Seat pricing charges a customer a fixed amount per named user or account per billing period. Usage-based pricing (also called consumption or metered pricing) charges based on how much of the product the customer actually consumes — in Buttondown's category, that would be metrics like subscriber counts or email volume rather than human logins.

Where each model works, and where it breaks.

Seat pricing has a well-documented advantage: revenue is predictable. A sales team can forecast next quarter from signed contracts, and finance can model growth without modeling customer behavior. It also aligns with buyers who budget annually. Its failure mode is that it punishes products where value does not scale with headcount. A team of three power users on a data-intensive product may extract enormous value while paying for three seats, while a 200-person team of casual users pays heavily for little usage. For small products, the sharper failure mode is different: per-seat prices that look trivial to enterprises can be the deciding factor for a two-person startup evaluating your tool.

Usage-based pricing inverts that tradeoff. Customers pay in proportion to value received, which lowers the barrier to adoption — a critical property for small products without a sales motion, where a self-serve signup must convert to paid with no human intervention. Metered pricing also lets a product monetize its heaviest users rather than capping them. The cost is unpredictability in both directions: customers cannot forecast their bills, which creates procurement friction and bill shock, and the vendor cannot forecast revenue, which complicates everything from hiring to runway math. Usage-based models also demand metering infrastructure — accurate, auditable tracking of the consumption metric — that a small team may not have built.

The metric choice is the product decision. For a vendor like Buttondown, the candidate consumption metrics are obvious: subscribers, emails sent, or some combination. The general principle generalizes: the metered unit must correlate with perceived customer value, be understandable before purchase, and be cheap to measure accurately. If the unit is opaque or gamed easily, usage pricing turns into a support burden rather than a growth lever.

Conditions to check before switching. Usage-based pricing tends to work when value concentrates in a minority of heavy users, when buyers are end users rather than procurement departments, and when the product can meter its value metric reliably. It tends to fail when customers demand flat, budgetable costs, when usage is spiky rather than steady, or when the metering unit diverges from how customers perceive value. Seat pricing holds up when teams adopt uniformly and headcount tracks with value; it leaks revenue when a few users capture most of the value.

Hybrid structures — a base platform fee plus usage overages — exist precisely to hedge these failure modes, trading some pricing simplicity for a floor on revenue and a cap on customer bill anxiety.

For small products specifically, the data-informed question Buttondown raises is the right one: pricing model selection is a distribution and unit-economics decision, not a fashion choice. As self-serve PLG motions continue to dominate how small products acquire customers, expect metered and hybrid pricing to keep spreading — and expect the winners to be the teams that instrument their value metric before they monetize it.