Consider an invented company with 600 employees, say a mid-size insurer’s technology arm in Thiruvananthapuram. The IT team is reviewing a new project-tracking SaaS tool that the business wants. Their first question is not about features. It is: “Does it support SSO?” To see why, follow two people through one Monday.

Monday, 9:30: Anjali joins

Anjali is a new analyst. IT creates her one company account in the identity provider, the central system that holds everyone’s login. Examples include Microsoft Entra ID, Okta and Google Workspace. She sets her password and registers a second factor, such as an authenticator app.

Now she opens the project tool. Without SSO, she would need a separate username and password there, and the same for every other SaaS product the company uses. With single sign-on, the tool sends her to the company’s login page. She proves who she is there, once, and the identity provider sends back a signed confirmation to the tool. The two usually speak a standard protocol, most often SAML or OpenID Connect. The tool never sees her company password. She is in within seconds, and the next application does not ask her again for the rest of the session.

Convenience is the visible benefit. The security benefit is bigger: login rules are enforced in one place. If the company requires a second factor, or blocks logins from unusual locations, or locks accounts after repeated failures, every connected tool inherits those rules without each vendor building them.

Monday, 11:00: does Anjali have the right access?

Logging in is one thing, but what she can do inside is another. The tool must know she belongs in the analytics team and may see those projects only. This is provisioning. Doing it by hand means someone in IT creates her account in the tool and sets permissions, one tool at a time. A widely supported standard called SCIM lets the identity provider create, update and remove accounts in connected tools automatically, based on her group in the directory. When HR records that she is in analytics, the tool follows.

Monday, 4:00: Ravi leaves

Ravi resigned and today is his last day. This is where SSO earns its keep. IT disables his single company account. Every application connected through SSO stops accepting him at once. Compare the manual world: he had accounts in perhaps a dozen tools, each with its own password. Some were created by teams who never told IT. A few months later, an auditor finds an active login belonging to someone who left. That is a real and common finding, and it is why security reviewers care. Where SCIM is in place, his accounts are also removed or deactivated in the tools, not merely blocked from logging in.

What the IT reviewer will actually ask you

  • Which protocols are supported (SAML, OpenID Connect), and which identity providers have you tested with?
  • Is automatic provisioning and deprovisioning supported, through SCIM or another method?
  • Can we force SSO for everyone and turn off local passwords, or can users still log in around it?
  • Does the tool support role mapping from our directory groups?
  • Is SSO included in the plan we are buying?

That last one is a sore point. Some vendors keep SSO and provisioning only in their highest tier, a practice customers sometimes call the “SSO tax”. Buyers dislike it, and it can derail a deal late in the process when the plan they priced turns out not to include it. Know what is in your own packaging before you quote.

Selling it honestly

Do not dismiss the request as an IT detail. For larger customers it is often a gate: without it, the tool does not get approved. Equally, do not claim support you have not verified. If your SSO works only with one identity provider, say so, and bring in a technical colleague early. Raising it yourself during discovery, before the security review, tells the buyer you understand their world, and it avoids a nasty surprise in week nine.

Figures and names in the examples are illustrative.

Leave a Reply

Trending