SaaS (Software-as-a-Service)

Software-as-a-Service (SaaS) is a software delivery model in which an application is hosted centrally by the vendor — usually on cloud infrastructure like AWS, GCP, or Azure — and accessed by customers over the internet, typically through a browser, in exchange for a recurring subscription fee. Unlike the old on-premise model, where a company bought a perpetual license and ran the software on its own servers, SaaS shifts hosting, security patching, uptime, and scaling responsibility to the vendor. This is why SaaS businesses obsess over metrics like MRR, churn, and CAC: the entire economic model depends on customers staying subscribed long enough to recoup the cost of acquiring them (see LTV\:CAC ratio). For builders, "SaaS" is now shorthand for a specific architecture pattern too: multi-tenant database design, centralized authentication (often via SSO/OAuth), a metered or seat-based billing engine (Stripe Billing, Paddle, Chargebee), and a single continuously-deployed codebase serving every customer simultaneously — as opposed to bespoke, per-customer installs. Concrete example: a three-person team building a scheduling tool doesn't ship a downloadable .exe; they deploy one Next.js app to Vercel, provision one Postgres database with a `tenant_id` column on every table, wire up Clerk for auth, and charge $29/month via Stripe. Every customer hits the same codebase and gets bug fixes and new features the moment they're deployed — no manual upgrade cycle. The SaaS model has expanded into vertical flavors (vertical SaaS, e.g. Toast for restaurants), infrastructure flavors (PaaS, IaaS), and API-first flavors (see API-first), but the throughline is always: centrally hosted, subscription-priced, continuously updated. SaaS pricing itself has evolved too — early SaaS was almost exclusively seat-based, but usage-based pricing has become standard for infrastructure and AI-heavy products where cost genuinely scales with consumption rather than headcount. Builders evaluating "should this be SaaS?" for a new product idea should weigh the tradeoff explicitly: SaaS wins on distribution speed, recurring revenue predictability, and zero-friction updates, but it also means the vendor inherits every operational burden — uptime, security, data protection, support — that an on-premise vendor could once hand off to the customer's own IT department. The category has also fragmented into recognizable sub-types worth knowing by name: B2B SaaS (sold to businesses, typically higher ACV and longer sales cycles) versus B2C SaaS (sold to consumers, typically lower price points and higher volume); horizontal SaaS (a general-purpose function like project management, usable across any industry) versus vertical SaaS (built for one industry's specific workflows); and increasingly "AI-native SaaS," where a language model is embedded directly into the core workflow rather than bolted on as a side feature.