The value metric

The value metric is the unit a SaaS product charges by - the thing that increases on the bill as the customer gets more value. Seats, active users, API calls, contacts stored, gigabytes, transactions processed, messages sent. It sounds like a pricing detail. It’s the most important pricing decision there is, because it sets whether revenue grows as the customer succeeds or stays flat while they do.

The test for a good value metric: as the customer gets more value from the product, do they naturally pay more, without you having to sell them anything? When the metric and the value move together, expansion is automatic. Slack charges per active user, so a company that rolls it out more widely pays more because it’s getting more, and nobody has to run an upsell. Stripe charges a slice of transaction volume, so as a merchant grows, Stripe grows with it. That alignment is why those models throw off net revenue retention above 100% almost mechanically.

Why this sits above the pricing page

Most pricing CRO is rearranging tiers, testing anchor prices, toggling the annual-default. All of that operates within a chosen value metric. If the metric is wrong, none of it matters, because you’re optimising how steeply price climbs on an axis that doesn’t track value.

The classic failure is the flat per-seat tool sold into a use case where seats don’t grow. The customer gets enormous value, uses it constantly, and pays the same every month because their seat count never changes. Revenue is capped by design, and no pricing-page test fixes that - the fix is changing what you meter.

Choosing the axis

A usable value metric has three properties:

• It tracks value the customer actually feels. Not an internal cost. Customers accept paying more for more contacts in a CRM, because more contacts is obviously more value. They resent paying per “API call” they can’t see or predict.
• It’s predictable enough to budget. A metric that spikes unexpectedly - pure usage-based billing with no ceiling - creates bill shock and churn. Some predictability, via tiers or caps, keeps the customer in control.
• It grows with the account. The whole point. A metric that’s flat for a growing customer leaves expansion revenue on the table.

Per-seat is popular because it’s predictable and easy to understand, but it only works when seats genuinely correlate with value. Usage-based aligns best with value but is the hardest to budget. The strongest models often combine a platform fee with a usage component, getting predictability and alignment at once.

Where it breaks

• Metering something the customer can’t control. If the bill moves on a number the user can’t influence, every invoice feels arbitrary and churn rises.
• Choosing a metric that punishes engagement. Charging per “login” or per “report run” teaches users to use the product less to save money, which is the exact opposite of what drives retention.
• Switching the value metric on existing customers. Sometimes necessary, always painful - it re-prices everyone’s relationship at once and triggers a churn event. Grandfathering softens it, but the migration is a project, not a test.