Usage-Based Pricing for SaaS: The PLG Practitioner's Guide

Usage-based pricing (UBP) charges customers for what they actually consume rather than a flat per-seat fee. Done well, it aligns your revenue with the value customers receive and makes expansion automatic. Done badly, it makes revenue unpredictable and confuses buyers. This guide covers the strategy, the models, and the implementation realities — from someone who has evaluated pricing transitions at scale.

What Is Usage-Based Pricing?

Usage-based pricing is a model where customers pay based on consumption — API calls, compute hours, transactions processed, credits burned — instead of a fixed fee per user per month.

The canonical examples:

• AWS charges for compute hours and storage consumed
• Stripe charges a percentage per transaction
• Twilio charges per API call and message sent
• Snowflake charges for compute credits consumed

The common thread: the customer's bill grows as they get more value from the product. There is no upsell conversation required for revenue to expand — usage is the expansion motion.

Usage-Based vs Seat-Based Pricing: Which Model Fits?

The 4 Main Usage-Based Pricing Models

1. Pay-as-you-go. Pure consumption, no commitment. Lowest friction to start, hardest to forecast. Best for developer tools and APIs where the first call should cost nothing.
2. Tiered usage. Bands of consumption at different unit rates — the first 10,000 events at one price, the next 90,000 cheaper. Rewards growth while keeping small customers affordable.
3. Hybrid. A seat or platform base fee plus usage overage. This is what most successful "usage-based" companies actually run — the base fee stabilizes revenue, the overage captures expansion.
4. Outcome-based. Pay per result achieved — per resolved ticket, per qualified lead, per funded loan. The purest value alignment, but requires airtight attribution and is hardest to implement.

How Usage-Based Pricing Impacts Revenue Predictability

This is the question every CFO asks, and the honest answer is: predictability moves, it doesn't disappear.

• NRR becomes the north star instead of MRR. In a seat model, revenue is contracted and stable. In a usage model, revenue is a trailing indicator of customer success. Net revenue retention — how much a cohort's spend grows or shrinks — replaces new MRR as the number that predicts the business.
• Cohort-based revenue modeling replaces flat ARR projection. You forecast by modeling how each signup cohort ramps its consumption over 3–12 months, not by counting signed contracts. Mature UBP companies forecast cohorts the way SaaS companies forecast pipeline.
• Land-and-expand is native to the model. Customers start small because starting small is free or cheap, then grow. Your "sales motion" is the onboarding experience.
• Seasonality and churn look different. Usage dips before customers churn — consumption decline is your earliest churn signal, often 60–90 days before a cancellation. Instrument it.

Transitioning From Seat-Based to Usage-Based

The transition is where most companies stall. The sequence that works:

• Grandfathering strategy. Existing customers keep their seat pricing for a defined period (typically 12 months) or until renewal. Forcing migration mid-contract is how you generate churn and support tickets simultaneously.
• Metering infrastructure. You cannot bill on usage you cannot measure. Metering must be accurate, auditable, and visible to the customer in near-real-time — a usage dashboard is a billing feature, not a nice-to-have.
• RevOps implications. Quoting, invoicing, revenue recognition, and comp plans all change. Sales comp in particular needs redesign — paying commission on signed seats doesn't work when revenue arrives over 12 months of consumption.
• Customer communication. Frame the transition around control: "pay for what you use" lands better than "we're changing how we bill you." Give customers a usage projection tool before you give them an invoice.

Key Metrics for Usage-Based SaaS Companies

• Consumption rate — the percentage of purchased usage actually consumed. Below 60% means customers will churn at renewal; above 90% means you're leaving expansion revenue unpriced.
• Expansion MRR from usage overage — the growth layer that replaces the upsell motion.
• Net revenue retention — target above 120% for a healthy UBP company. Below 100% in a usage model is a red flag that consumption is shrinking.
• CAC payback period — usage models often have slower initial revenue per customer, so payback discipline matters more, not less.
• Usage-to-paid conversion rate — for freemium and PLG hybrids, the percentage of free-tier users whose consumption crosses the paid threshold.

Usage-Based Pricing and PLG

Usage-based pricing and product-led growth are natural partners. PLG says the product drives acquisition, activation, and expansion. UBP says revenue follows usage. Put them together and the loop closes itself: the product drives adoption, adoption drives consumption, consumption drives revenue — with no sales motion required for growth.

This is why the canonical PLG companies of the infrastructure era — Snowflake, Twilio, Stripe, AWS — are all usage-based. The pricing model is the growth model. If you're building a PLG motion on top of seat-based pricing, you're asking a sales artifact to do a product's job.

FAQs

What is usage-based pricing?

Usage-based pricing is a billing model where customers pay based on how much of the product they consume — API calls, compute, transactions, credits — rather than a flat per-seat subscription fee. Revenue scales automatically with the value the customer receives.

How does usage-based pricing work for SaaS?

The product meters consumption in near-real-time, applies a rate (flat, tiered, or hybrid with a base fee), and bills in arrears or against pre-purchased credits. Customers typically see a live usage dashboard so there are no invoice surprises.

What is the difference between seat-based and usage-based pricing?

Seat-based pricing charges a fixed fee per user per month, giving high revenue predictability but weak alignment with customer value. Usage-based pricing charges for consumption, giving strong value alignment and automatic expansion, but variable revenue that requires cohort-based forecasting.

How do I transition from seat-based to usage-based pricing?

Grandfather existing customers for a defined period, build accurate and customer-visible metering first, redesign RevOps (quoting, invoicing, comp) for consumption revenue, and communicate the change as customer control over spend rather than a billing change.

How does usage-based pricing affect revenue predictability?

Predictability shifts from contracted MRR to cohort-based consumption modeling. NRR becomes the north star metric, consumption decline becomes the earliest churn signal, and forecasting models how each cohort ramps usage over 3–12 months rather than counting signed contracts.

What are examples of usage-based pricing?

AWS (compute hours), Stripe (percentage per transaction), Twilio (per API call), and Snowflake (compute credits) are the canonical examples. Most successful implementations are actually hybrids — a base platform fee plus usage overage.

Is usage-based pricing better for PLG companies?

Usually, yes. PLG relies on the product driving acquisition and expansion without a sales motion, and usage-based pricing makes revenue expansion a direct function of product adoption. The pricing model and the growth model reinforce each other.

What metrics matter most for usage-based SaaS?

Consumption rate, expansion MRR from overage, net revenue retention (target above 120%), CAC payback period, and usage-to-paid conversion for freemium hybrids. Consumption decline is also your earliest churn warning.

---

Evaluating a pricing model change? Let's map the transition.