Per-seat, usage, or platform fee: choosing the right SaaS pricing meter.

We have argued in pricing power in vertical SaaS that narrow workflows earn the right to raise prices annually, and in how we migrate pricing after a B2B SaaS acquisition about the four-quarter sequencing we use to capture unrealized economics. This essay sits one layer up: it is the question of which meter the brand should be running before the migration begins.

The three meters

Almost every B2B SaaS pricing model is one of three meters, or a hybrid of two of them. Per-seat charges for access. Usage-based charges for an event the product processed. Platform-fee charges for a flat license, often paired with a usage tail. Each meter encodes a different theory of value, and the theory is more important than the price point.

The meter is a statement about who the product is for, what outcome the product is producing, and how the customer will measure their own success with the product. Get the meter wrong and every other pricing decision is downstream of a mistake.

Per-seat: the most overused meter

Per-seat pricing is the right meter when the value of the product is proportional to the number of humans using it, and the value of an additional human is roughly equal to the value of the previous one. Collaboration tools fit. CRM fits. Most help-desk and ticketing tools fit. The meter is honest because the value scales with the denominator.

Per-seat is the wrong meter when the product is doing work on behalf of the customer rather than enabling the customer to do work themselves. An AI-native workflow product that processes a million events for a team of five is creating value that has nothing to do with the five seats. Every dollar of price tied to seats is leaving money on the table, and the customer eventually notices when their volume grows ten-x without a price change. The meter that looks fair at signing is the meter that looks underpriced at renewal.

Usage-based: the most over-romanticized meter

Usage-based pricing is the right meter when the customer has a clear sense of what an event is worth to them and when the volume of events is reasonably predictable. The canonical examples are infrastructure, payments, and communications APIs, where the customer's own business model creates a natural ceiling on usage and a comprehensible per-event value.

Usage-based is the wrong meter when the customer cannot forecast their own volume and the bill becomes a source of monthly anxiety. It is also the wrong meter when the usage event is something the product is producing on behalf of the customer (a generated document, a classification, a recommendation) and the customer cannot easily verify whether the event was useful. The unpredictability and the verifiability problems combine to make usage-based the most-churned pricing model in the industry when it is misapplied, even though it is the most-modeled pricing model in board decks.

Platform-fee: the underrated meter

Platform-fee pricing — a flat annual license, often paired with a small overage on the high end — is the right meter when the brand is genuinely a system of record and the customer needs budget predictability. Vertical SaaS targeting industries with annual procurement cycles fits this model unusually well, because the buyer's planning cycle aligns to the meter and the procurement-finance relationship rewards predictability over precision.

Platform fee is the wrong meter when the brand has not actually earned system-of-record posture. Charging a platform fee for what the customer perceives as a feature is the fastest way to invite a budget conversation the brand will lose. The system-of-record posture is the precondition for the meter, not the marketing claim made on top of it.

How to decide whether the current meter still fits

The diagnostic question is simple, and it is rarely run cleanly. We ask the operating team three things on every brand in the first quarter after close.

• What is the unit of value the customer would name if asked? If the answer is different from the meter on the invoice, the meter is wrong.
• Where does the brand leave money on the table today? The most common answer is the high-volume long tail of the customer base where the seat count has not grown but the workflow volume has ten-x'd.
• What would happen to renewal if the meter changed? If the answer is that nothing structural changes for the customer, the meter is safe to migrate. If the answer is that the customer would re-evaluate the relationship, the migration is the wrong battle to fight in the first year after close.

The Cobalt Glacier remap pattern

The remap we do most often is not a wholesale swap. It is adding a usage component on top of an existing per-seat model, with the usage component priced low enough that the median customer's bill does not change in the first year. The usage component captures the long tail of high-volume customers immediately, sets up a multi-year price story that does not require an annual list-price fight, and gives the brand a meter that maps to the value the AI layer is now creating. The pattern is gentle on the customer base and effective on the economics, which is the rare combination pricing migrations are supposed to produce but usually do not.

The common mistakes

The most common mistake is over-engineering the meter. Pricing models with four dimensions, eight tiers, and a usage component on each tier are unsellable in the field and unenforceable in customer success. The meter has to fit on the back of a business card. Anything more complicated than that is a signal that the brand has not decided what the value actually is.

The second most common mistake is changing the meter and the price at the same time. The customer cannot tell which change is hurting them, the renewal motion gets contentious, and the brand learns nothing about the elasticity of either change. We change one variable at a time, deliberately, with the renewal calendar in mind.

Why the meter conversation usually surfaces something else

The meter conversation is rarely just about pricing. In almost every case, it surfaces a deeper question about what the brand is actually selling. The team that thought it was selling a tool to a five-person team discovers it is selling a workflow engine that processes a million events a month. The team that thought it was selling a system of record discovers it is selling a feature inside a larger workflow the customer already owns. The team that thought it had pricing power discovers it has been competing on convenience for a decade. Each of these realizations is uncomfortable in the moment and clarifying for the long arc. The pricing meter is the cheapest place to surface them because the meter is something the brand can change in a year without a product rebuild. The strategy questions the meter surfaces are usually the ones that needed to be on the table anyway, and the meter conversation is the forcing function that puts them there.

The bottom line

The pricing meter is the most strategic decision in a B2B SaaS business and the one most rarely revisited. Per-seat is the right meter when value scales with humans. Usage-based is the right meter when value scales with events the customer can forecast. Platform-fee is the right meter when the brand is genuinely a system of record. Most acquired brands are running a meter that no longer maps to their value, and the remap is one of the most leveraged moves we make in the first four quarters after close.

If you are a founder thinking about whether your current pricing model would survive a permanent-capital underwriting, read how we work with founders before and after close. The pricing meter conversation is one we have on every deal we evaluate, and the answer is rarely the one the founder expected when the diligence started.