Billing System Software: How It Actually Works

For a scaling SaaS startup, an inflexible billing system can directly inhibit growth. A company planning to shift from a $49/month flat fee to a usage-based model with three tiers and overage charges for API calls can find its legacy billing infrastructure incapable of handling the change. This forces a costly re-engineering project, delaying revenue and frustrating product teams. A billing system is product infrastructure. It needs to be as agile as what it supports.

Lago is the open-source billing engine built for this problem. Event-driven, API-first, and designed to handle metering, pricing logic, and invoicing without a dedicated billing team.

What is a SaaS billing system?

SaaS billing system software manages the entire financial lifecycle of a customer relationship in a recurring revenue model. Unlike traditional invoicing software, a SaaS-native system automates subscription management, recurring invoice generation, payment collection, and revenue recognition. These systems are built for the high-volume, variable nature of SaaS pricing: recurring fees, one-time charges, and metered usage, often all at once.

By automating these workflows, businesses accelerate time-to-cash, reduce payment failures through automated dunning, and minimize the revenue leakage caused by manual billing errors. The core function is translating product usage and subscription terms into accurate, timely invoices and recognized revenue. That directly impacts MRR and NRR.

The architecture

A billing system is composed of several layers. They need to work together.

The core is the metering engine: it ingests raw usage events, aggregates them by customer and period, and applies the pricing rules defined in each plan. A company charging per gigabyte of storage or per 1,000 API calls needs this to be accurate and real-time. Errors here compound directly into billing disputes.

Invoicing sits downstream from metering. The billing system generates invoices from subscription and usage data, sends them via email or customer portal, and handles the edge cases: pro-rations, credits, multi-currency, multi-entity.

Payment processing connects to gateways (Stripe, Adyen, GoCardless) via API. Multiple payment methods (credit card, ACH, SEPA) and multi-currency support are table stakes for any company selling internationally.

Dunning is where involuntary churn gets recovered or lost. A properly configured retry schedule with customer-facing communication can recover a significant portion of failed payments that would otherwise lapse. Most companies underinvest here.

Revenue analytics is the reporting layer: MRR, ARR, churn, NRR, LTV. Critically, the system must also handle revenue recognition compliance under ASC 606 and IFRS 15. Revenue recognized as earned, not as billed.

Pricing models

The simplest model: a flat monthly or annual subscription. Predictable for both sides. Breaks when customer value diverges significantly from the flat fee. A 10x-usage customer pays the same as a light user.

Per-seat pricing scales with headcount. Clean to explain, easy to forecast. Breaks when users share credentials or when the product is infrastructure with no meaningful "user" concept.

Usage-based billing charges on actual consumption: API calls, tokens, GB stored, minutes processed. Revenue scales with customer value. Forecasting is harder. Metering must be accurate and real-time.

Hybrid pricing combines a flat platform fee with usage-based overages. A base fee covers a usage allowance; anything above is metered. Most AI and API products use this structure. It gives customers cost predictability and gives vendors protected base revenue.

Graduated pricing charges different unit prices by tier: first 100 units at $1.00, units 101-500 at $0.80, units 501+ at $0.60. Each unit is priced at the tier it falls in. Volume pricing applies one rate to the entire quantity once a threshold is reached. Different math, different incentive structure.

Implementing complex models, especially usage-based and hybrid, requires a system with flexible pricing logic. If every pricing change requires engineering work, shipping new models takes weeks instead of hours.

Build vs. buy

Building in-house offers complete control but comes at real cost. The initial engineering investment is substantial. The ongoing maintenance burden (security, global tax regulations, new pricing models) diverts resources away from the core product. What starts as a simple script to charge credit cards can quickly evolve into a complex, brittle system that slows everything down.

Buying a dedicated billing platform accelerates time-to-market and reduces total cost of ownership. Engineering teams focus on building features customers value, not reinventing billing infrastructure.

Lago is open-source and can be self-hosted in your own VPC, on-premise, or air-gapped. This is not a niche option for edge cases. It's for any company that wants billing infrastructure they own: no vendor lock-in, full visibility into the code, deployable anywhere. The alternative to building billing from scratch isn't only a cloud SaaS. It's open-source infrastructure you run yourself.

How billing system APIs work

Modern billing platforms expose a RESTful API that lets developers manage the entire billing lifecycle programmatically. Key resources include Customers, Subscriptions, Plans, and Invoices. Operations run through standard HTTP methods (GET, POST, PUT, DELETE). Webhooks push real-time notifications based on events like invoice.paid, subscription.canceled, or payment.failed. These events enable an event-driven architecture that automates downstream workflows without polling.

Integration points

A billing system sits at the center of the commercial stack. The product application sends usage events to the billing system's metering endpoint in real time. Customer and subscription data syncs with the CRM (Salesforce, HubSpot) so sales and support have a current account view. Invoices, payments, and revenue recognition schedules push to the accounting system (NetSuite, QuickBooks) for financial reporting and reconciliation.

Security and compliance

The payment processor (Stripe, Adyen, GoCardless) handles PCI DSS compliance for cardholder data. The billing engine connects to it via API but doesn't store card data itself. On the billing infrastructure side, Lago is SOC 2 Type II certified and undergoes regular third-party penetration testing. SSO, RBAC, and audit logs are available for enterprise access control requirements. Startups can sell into enterprise accounts without becoming compliance experts. But the distinction between the payment processor's certifications and the billing engine's certifications matters.

Built by Lago, open-source billing infrastructure for usage-based and hybrid pricing. Explore the docs or book a demo.