Usage-Based Pricing: The First Principles That Determine Whether It Works

Usage-based pricing sounds like a principle. Pay for what you use. Align cost with value. Reward growth. These statements feel self-evidently correct, which is why so many teams adopt consumption models without questioning the underlying logic.

When you strip usage-based pricing down to its first principles, you find that three conditions must be true for the model to work. Most implementations violate at least one of them.

What You're Paying For It

The cost of a first-principles failure in usage-based pricing is not just revenue leakage. It is customer relationship damage that is slow to build and difficult to repair.

When the billing metric drifts from perceived value, customers develop a baseline suspicion of your invoices. They audit usage reports, challenge charges, and route renewal conversations through procurement rather than the champions who bought the product. An account that once expanded automatically now becomes a negotiation every renewal cycle. At $25M annual recurring revenue (ARR) with 40 percent of revenue in usage-based contracts, converting 20 percent of those accounts from automatic renewal to negotiated renewal adds 80 to 120 additional hours of customer success (CS) and sales time annually, before counting the annual contract value (ACV) impact of negotiated decreases.

The Operating Play

Principle 1: The billing metric must be observable by the customer without engineering support.

If your customer needs to ask your support team to explain their invoice, your billing metric is wrong. The metric must map to something the customer tracks independently. A task management product billing per task completion works because customers track completions as a business outcome. The same product billing per API call fails because customers do not track API calls as a business outcome.

Test your current metric: send your 5 largest accounts a usage report. Ask them to verify it against their own records without help from your team. The percentage who can do this accurately is a direct measure of metric transparency.

Principle 2: Usage must scale with customer success, not with customer inefficiency.

A storage product billing per gigabyte stored rewards data hoarding and penalizes customers who organize their data well. A communication platform billing per message sent rewards verbosity and penalizes efficient teams. Usage metrics that scale with inefficiency create perverse incentives that undermine product adoption.

The correct framing is: does this metric go up when the customer achieves more of their desired outcome? If usage rises because the customer is succeeding more, the metric is aligned. If usage can rise for reasons unrelated to success or even inversely correlated with success, the metric is misaligned.

Principle 3: The customer must be able to predict their bill before committing to an annual contract.

Unpredictable bills are a structural barrier to enterprise sales. Procurement teams cannot approve open-ended consumption commitments. If your usage-based model does not have a credible annual spend forecasting method, your sales motion will default to quoting annual minimums that leave expansion upside uncaptured and underestimate true contract value.

Build the forecasting model from your existing customer data. Show prospects what accounts of their size and use case type typically spend over 12 months, with a range. That range gives procurement a ceiling to budget against.

The Hidden Failure

A $14M ARR document processing platform chose "pages processed" as their billing metric. On first-principles review, this metric fails all three tests. Customers do not independently track pages processed. Processing more pages may indicate a customer is re-running failed jobs, not succeeding. And pages processed is highly variable based on document complexity, making annual forecasting unreliable.

The company spent 2 years on this model before switching to "documents completed without re-submission." This metric is observable (customers track completed documents as business outcomes), scales with success (more completions means more value), and is forecastable from document volume estimates. Net revenue retention (NRR) improved from 98 percent to 114 percent within 4 quarters of the transition.

The new metric was available from day one. Nobody had applied first-principles thinking to the original choice.

Start Here This Week

Write down your current billing metric and test it against all three principles above. Be honest about which ones fail. You do not need to change the model this week, but you need to know which assumptions are currently unexamined.

Book a 15-minute working session at Assess Your Pricing Health to run this analysis with a pricing specialist.

For related reading, see the hypothesis-led approach to usage-based pricing and the operator's guide to usage-based pricing.