Skip to main content

Single Sign-On (SSO) - One Login for Everything

About 2 min read

Single Sign-On (SSO) is a mechanism that lets you access multiple services and applications with a single authentication. Users no longer need to enter a password for each service, which greatly reduces password fatigue. In corporate IT environments the number of cloud services used for work is large, so keeping a separate strong password for every single service is not realistic. For the technical details, also refer to the SSO term entry.

The Relationship Between IdP and SP

At the core of single sign-on is the trust relationship between the IdP (Identity Provider) and the SP (Service Provider). The IdP centrally manages user authentication, and the SP verifies the authentication token issued by the IdP to grant access. Representative IdPs include Okta, Microsoft Entra ID (formerly Azure AD), Google Workspace, and OneLogin.

User accesses the SP
SP redirects to the IdP
Authenticate at the IdP (only once)
Token issued
Access to all SPs

With this mechanism, users can seamlessly use all linked services such as Slack, Salesforce, Jira, and Confluence by logging in to the IdP just once.

SAML vs OIDC Comparison

SAML 2.0 and OpenID Connect (OIDC) are the protocols widely used to implement single sign-on. Which one to choose depends on the use case.

ItemSAML 2.0OpenID Connect (OIDC)
Data formatXMLJSON (JWT)
Main useEnterprise SaaS integrationWeb / mobile apps, social login
Base protocolProprietary specificationExtension of OAuth 2.0
Implementation complexityHigh (XML signature verification is complex)Low (REST API based)
Mobile supportPoor (assumes browser redirects)Good (supports native apps)
Adoption examplesSalesforce、Workday、ServiceNowGoogle、GitHub、Auth0

SAML still dominates for existing enterprise SaaS integrations, but more new development is adopting OIDC. OIDC is favored because it manages session tokens in a JSON-based, easy-to-handle way, lowering the implementation barrier for developers.

Risks of Single Sign-On

Single sign-on offers significant benefits for both convenience and security, but it also carries inherent risks.

Single point of failure

If the IdP goes down, you can no longer log in to any linked service. In the 2024 Okta outage, many companies were forced to halt operations.

Risk of credential concentration

If the IdP account is compromised, unauthorized access to all linked services becomes possible. The IdP account requires the strongest protection.

Complexity of session management

When the session lifetimes of the IdP and each SP differ, logout inconsistencies arise. Implementing single logout (SLO) is technically difficult.

To mitigate these risks, always enable multi-factor authentication on the IdP account and, where possible, adopt phishing-resistant MFA using passkeys or security keys. It is also important to prepare an emergency access method (a break-glass account) in advance for IdP outages. Corporate password policy and startup security checklist explain organization-level adoption guidelines.

The Effect of Relieving Password Fatigue

The greatest practical benefit of single sign-on is relieving password fatigue. The number of passwords employees manage drops dramatically, curbing dangerous behaviors such as password reuse and simplification. "I forgot my password" and "my account is locked" are staple tickets for any IT help desk, and consolidating the login into a single entry point shrinks the pool of requests itself. By integrating with an IAM platform, account management at onboarding and offboarding can be centralized, resolving the "zombie account" problem in which departing employees retain access rights. Understanding the risks of OAuth permissions as well will help in designing the overall authentication infrastructure.

Real-World Use Cases

Consider a company whose internal SaaS count has grown into the teens. A common approach is to settle on a single identity provider (IdP) first, then connect the most heavily used tools one at a time, starting with chat, document sharing, issue tracking, and source control. Switching everything over at once inevitably hits a workflow that cannot be paused, so it is more practical to rehearse the procedure on a low-impact service before expanding. The first visible payoff is help desk load: password reset requests decline, and deactivating a departing employee can be completed from the IdP alone. The flip side is that an IdP outage stops every login at once, so the standard practice is to keep an emergency administrator login path outside single sign-on.

Related Terms

Was this article helpful?