Article
Usage-based billing, explained
AI is changing how software businesses build, price, and serve customers. Your financial infrastructure needs to keep pace.
Pricing has never been simple. For years, though, SaaS businesses had a reliable default: seat-based pricing, which spreads the high cost of development across many users.
While this model worked for years, generative AI is rewriting the story. Today, the cost of serving each user depends on how much they use AI inside your product, and many software businesses are finding that seat-based pricing no longer covers the cost of delivery.
That’s where usage-based pricing comes in. Instead of charging for access, it charges for consumption. That keeps pricing aligned with the cost of delivery and makes it more transparent for customers.
Usage-based billing is growing in popularity, but it comes with plenty of questions. Keep reading to learn more about:
What is usage-based billing
Seat-based vs usage-based billing
Usage-based billing models
How to automate usage-based billing
What is usage-based billing?
Usage-based pricing charges customers for what they consume, not what they can access. Instead of a flat rate per person per month, you pick a unit that represents the work your product does, count it, and bill on it. It's also known as consumption pricing.
Seat-based vs usage-based billing
SaaS businesses have traditionally used a seat-based pricing model, which charges customers per license to the platform. This model worked well because the cost per user was stable. In most cases, the price per seat was a fair proxy for the software’s value to each user.
However, as more companies add AI features to their products, underlying cost structures are becoming more complex.
AI makes your product more valuable: A single power user can now get more done inside your platform with AI tools and agents.
AI costs vary with usage: Computing costs can differ widely from one user to the next, so the cost of delivery per user is no longer predictable.
As a result, seat-based pricing no longer reflects the value each user gets from your product, or covers the cost of delivering it.
That’s why usage-based billing is becoming increasingly popular. It keeps revenue aligned with operating costs, which protects your margins as customers adopt new features and workflows.
What are the benefits of usage-based billing?
Many AI-native companies already use usage-based billing for three main reasons:
Usage-based billing grows revenue with usage
Seat-based pricing takes a one-size-fits-all approach. Usage-based billing lets you earn more from the customers who get the most out of your product, so a power user contributes more revenue than someone who logs in once a month to check a dashboard. It also shows which features and products generate the most revenue, which can help shape your roadmap.
Usage-based billing protects your margins as you scale
Usage-based billing aligns your product's pricing with the computing costs each customer generates. As customers explore new workflows and features, your pricing automatically reflects their growth and adoption without requiring constant contract renegotiation.
Usage-based billing makes software pricing more transparent
Seat-based billing is predictable, but it isn’t transparent. With usage-based billing models, customers pay for what they use, not for the value the product is assumed to create. Many customers find this a closer match to the value they get from the tool. It can also lower the barrier to trying new features, which can increase adoption.
Examples of companies that have switched to usage-based billing
A growing number of enterprise software businesses are moving away from per-seat pricing. Some that have switched to usage-based billing include:
Jasper launched usage-based monetization for its most powerful AI tools in just two months, without diverting engineering resources away from product development.
Vercel used usage-based billing tools to reduce manual reconciliation and launch billing for a new product in three weeks.
Supabase switched to usage-based billing and now processes 1.5 million invoices per month across different pricing structures and models.
What are the main usage-based billing models?
Many companies opt for hybrid pricing models that mix subscriptions with usage-based elements. These models offer predictable recurring revenue, but still protect margins as computing costs vary.
Common hybrid structures include:
Subscription + usage overage: The customer pays a monthly fee that includes a certain amount of usage. Any usage above that threshold incurs overage charges. These usage-based recurring billing subscriptions are where most enterprise companies settle.
Subscription + tiered limits: The customer pays a monthly fee that includes a certain amount of usage. When they hit that limit, they’re prompted to upgrade to the next tier.
Freemium + pay-as-you-go: The customer receives a free monthly usage limit, then is charged for usage above that limit.
User seat + usage: The customer pays a per-seat fee for access, plus an additional charge for AI usage.
Subscription + marketplace/add-ons: The customer pays a subscription fee for core use, then buys add-on features or models at an additional rate.
How to set up usage-based billing for SaaS
Implementing usage-based billing automation is easier said than done. There are two things to consider as you set it up:
How to automate usage-based billing: Metering, rating, and invoicing all need to run without manual intervention.
Revenue recognition for usage-based billing needs adjustment: Your existing sales structure and revenue function may require organizational shifts.
Many subscription billing providers have attempted to add usage-based capabilities to their platforms. But these tools are usually built for static pricing, which makes dynamic pricing strategies hard to execute.
That’s why usage-based billing works best on a purpose-built platform that brings together invoicing, reporting, revenue analysis, revenue simulation, and, increasingly, payments in one solution.
How to manage usage-based billing
Whether you’re using usage-based billing for fintech or usage-based billing for AI companies, your goal is to keep costs aligned with revenue. To establish your strategy, consider the following steps:
1: Choose a cost-related pricing metric
When establishing what to measure and price, you need to ensure your chosen metric correlates with your cost of delivery. You also want to pick a metric that aligns with how customers use your product.
Common units include:
Per query
Per 1k tokens processed
Per image generated
Per conversation
2: Establish usage tiers and limits
The next step is to define your packages and subscriptions. Most companies choose a mix of fixed subscriptions, usage fees, and add-ons. For high-usage plans, it may be worth creating custom quotes.
3: Monitor your cost-to-revenue ratios
Compare your costs against revenue per customer or per usage unit. You can adjust your pricing as needed to maintain higher margins across all customers.
Where to start
Usage-based billing isn't a switch you flip at renewal. It changes how you price, how you forecast, how you compensate sales, and what customers see on an invoice.
Implementing generative AI usage-based billing software can be complex, so it pays to work with experts as you design your strategy and roll out your technology. Done well, it can strengthen your relationship with customers, protect your margins, and set you up for increased profitability as you add more AI tools and services.
FAQs
Your product records every billable action as it happens, and the billing system counts those records. If you charge per API call, every call gets logged with a timestamp, the customer it belongs to, and anything that affects the price, like which model ran or which region it ran in. Get that log right and you can change how you charge later without touching the product again.