Home / Wiki / S / Software as a Service (SaaS)

Software as a Service (SaaS)

Technology and VendorsUpdated 16 Nov 2026

Software as a service (SaaS) is a way of delivering software in which the provider hosts the application and customers use it over the internet, normally for a recurring fee. It replaced the older model in which a customer bought a licence, installed the program on its own machines and paid separately for maintenance and upgrades.

SaaS is one of the main layers of cloud computing, alongside rented infrastructure and development platforms. It has become the standard form of business software for tasks such as customer management, accounting, payroll, collaboration and payments. For investors it is a distinct category of company, analysed with its own set of measures and valued chiefly on recurring revenue.

01

How the model works

A SaaS provider runs one version of its application for all customers, usually on infrastructure rented from a hyperscaler, and takes responsibility for updates, security and availability. Customers reach it through a browser or an app. They pay by subscription, most often a price per user per month, in tiers with different features, or through usage-based billing tied to volume. Contracts run monthly, annually or for several years. For the customer, the spending is an operating cost instead of capital expenditure.

02

History

Shared access to remote computers dates from the time-sharing services of the 1960s. In the late 1990s, application service providers hosted conventional software for clients, with limited success. Salesforce, founded in 1999, sold customer management software built from the start for delivery through a browser and became the model for the industry. Faster internet connections and more capable browsers widened what could be delivered this way, and during the 2010s the established software houses moved their products from one-off licences to subscriptions.

03

Economics and key metrics

A SaaS company spends heavily to win a customer and then collects revenue over the life of the relationship. Analysts therefore focus on a small group of measures.

  • Recurring revenue. Annual recurring revenue is the yearly value of the subscriptions in force.
  • Acquisition cost. Customer acquisition cost is compared with the revenue a customer brings, often as the number of months needed to recover it.
  • Profitability. Gross margin is high because serving one more user costs little. The Rule of 40 adds the growth rate to a profit margin as a rough test of balance.
  • Retention. Churn is the revenue lost from customers who leave or downgrade. Net revenue retention shows what a group of existing customers pays one year later.

Net revenue retention is calculated as:

NRR=A+E−C−LA\text{NRR} = \frac{A + E - C - L}{A}
  • AA is the recurring revenue from a group of customers at the start of the period.
  • EE is expansion: extra revenue from those customers through more users, higher tiers or price rises.
  • CC is contraction, the revenue lost through downgrades.
  • LL is the revenue lost from customers who cancel.

If a group of customers paid $10.0 million a year at the start, added $1.5 million, downgraded by $0.3 million and cancelled $0.7 million, the result is (10.0+1.5−0.3−0.7)/10.0=1.05(10.0 + 1.5 - 0.3 - 0.7) / 10.0 = 1.05, or 105 percent. A figure above 100 percent means the existing customer base grows without any new sales.

04

Accounting and valuation

Under current revenue recognition standards, subscription fees are recorded evenly over the term of the contract. Cash received in advance is carried as deferred revenue, and contracted amounts not yet billed are disclosed as a revenue backlog. Because selling costs are incurred up front and revenue arrives later, a fast-growing company can report losses while its customer relationships are profitable. Share awards to staff are a large expense in the sector, measured by the stock-based compensation burden.

SaaS companies are commonly valued as a multiple of revenue, using enterprise value or the price-to-sales ratio, since many have little or no profit. Such valuations depend on cash flows far in the future and are sensitive to interest rates: multiples reached a peak in 2021 and fell sharply in 2022 as rates rose.

05

Risks and the effect of artificial intelligence

The general risks are customer losses, competition from cheaper rivals, security breaches, outages and dependence on a few infrastructure providers. Since 2023, attention has turned to generative AI. One concern is that an AI agent able to carry out tasks could reduce the number of human users an application needs, which would undermine pricing per user. Another is that customers or new entrants could build substitutes more cheaply. Running AI models also raises the cost of delivering a service, which lowers gross margin.

Providers have responded by adding AI features to their products and by testing prices based on usage or results. Whether AI weakens established SaaS companies or strengthens those that hold customers' data and workflows is unsettled, and views among analysts differ widely.

Last updated 16 Nov 2026