Skip to content

New: How We Design Technology - our seven-layer architecture model, explained in full. See the architecture

general

Identity-First Security: Why Identity Is Your IT Foundation

11/05/2026

A regional sales manager opens the CRM from a hotel business center. A contractor whose project wrapped up last month still has a working login to the shared project site. A finance clerk approves a payment from a personal tablet. None of these sessions crosses your office firewall, and every one of them comes down to the same question: does your identity system believe this person, on this device, should get in?

That is why identity-first security has moved from a slogan to a design principle. When most of your applications live in the cloud and your people work from anywhere, the sign-in is where access gets granted or refused. Designing identity first, and treating it as its own deliberate layer, gives you stronger control today and keeps tomorrow's vendor decisions open.

At a glance

  • Cloud apps and remote work moved the access decision from the network edge to the sign-in, so identity now carries most of the security load.
  • A strong identity layer means one authoritative provider, phishing-resistant MFA, device-aware access policies, disciplined account lifecycles, and protected emergency access.
  • Keeping identity separate from your email and collaboration vendor lets you change either one later without rebuilding the other.

Why identity now decides who gets in

For years, the network did most of the gatekeeping. If you were inside the building or on the VPN, you were largely trusted. That model weakens when email, files, CRM, payroll, and line-of-business apps run as SaaS services reachable from any internet connection. An attacker with a valid username, password, and an approved MFA prompt doesn't need to touch your network at all.

NIST describes this shift in its Zero Trust Architecture publication. Zero trust moves defenses from static, network-based perimeters toward users, assets, and resources, and it grants no implicit trust to an account or device based on its physical or network location. Authentication and authorization happen before a session to a resource is established. In practice, that puts your identity provider in the middle of almost every access decision your organization makes.

The phrase "identity is the new perimeter" gets repeated so often it can sound like marketing. The underlying point holds up. If identity is weak, every other control is working with bad information.

What the frameworks say, at a high level

Two public references are worth knowing, even if you never pursue a formal zero trust program.

NIST SP 800-207 defines the concepts and components of a zero trust architecture. It is vendor-neutral and describes what the architecture needs to do, which makes it useful for checking whether a product pitch matches a real requirement.

CISA's Zero Trust Maturity Model (version 2.0, released in April 2023) organizes the work into five pillars: identity, devices, networks, applications and workloads, and data. It describes four maturity stages: Traditional, Initial, Advanced, and Optimal. In the identity pillar, the Advanced stage describes authenticating all identities with phishing-resistant MFA, and the Optimal stage describes validating identity continuously, well past the moment access is first granted.

Neither framework is a certification, and neither tells a 200-person organization exactly what to buy. They are most useful as a shared vocabulary and as a way to show leadership where you are today and what the next stage would require.

Designing identity-first security in practice

The practical work fits into five design decisions. Most organizations already have pieces of each one. The gaps usually show up where they meet.

1. Choose one authoritative identity provider

Every important application should trust a single identity provider for sign-in, through single sign-on standards such as SAML or OpenID Connect where the app supports them. When some apps keep their own local passwords, offboarding becomes a scavenger hunt and MFA coverage has holes. Pick the provider that will be the source of truth, document which systems feed it (HR, an on-premises directory, or both), and keep a list of the apps that still sit outside it.

2. Move to phishing-resistant MFA, starting with the accounts attackers want most

Push notifications, text codes, and one-time passcodes are far better than passwords alone, and they can still be phished or approved by a tired user. Phishing-resistant methods, such as FIDO2 security keys and passkeys, Windows Hello for Business, and certificate-based authentication, tie the sign-in to the legitimate site or device. Start with administrators, finance staff, executives, and anyone who can approve payments or change access. For how this change feels to the people using it, see passwordless authentication.

3. Tie access to device health

A correct password on an unmanaged laptop carries more risk than the same sign-in from a company device with current patches and disk encryption. Conditional access policies can require a device to meet your compliance rules before it reaches sensitive data. In a Microsoft environment, that means Intune compliance policies feeding Microsoft Entra ID Conditional Access, which requires Entra ID P1 licensing. Okta and other providers offer comparable device signals through their own integrations. Roll these policies out in report-only mode first so you can see who would be blocked before anyone actually is.

4. Run a real joiner, mover, leaver process

Most identity problems are lifecycle problems. A new hire gets access by copying a coworker's groups. Someone changes roles and keeps the old permissions. A contractor's account outlives the contract. Connect account creation and removal to an authoritative source, usually HR, and assign access by role so changes follow the person. Review privileged roles and external guest access on a fixed schedule.

5. Govern service accounts and emergency access

Service accounts and integrations often hold broad permissions and never sign in interactively, so they slip past user-focused policies. Inventory them, give each one an owner, and limit what each can reach. Separately, keep a small number of emergency access (break-glass) accounts that can still get in if a conditional access policy or MFA service locks everyone out. Microsoft's conditional access guidance recommends excluding these accounts from policies that could cause a lockout. Protect them with strong credentials, store those credentials securely, alert on every use, and test them on a schedule.

Why identity should be its own layer

Many organizations never ask one architectural question out loud: is your identity provider a deliberate choice, or did it arrive bundled with your email?

For many organizations, Microsoft Entra ID showed up with Microsoft 365, or Google's identity services came with Google Workspace. That can work well. The risk is that identity, device management, collaboration, and security policy become so intertwined that changing any one of them turns into a rebuild of all of them. If a price increase, an acquisition, or a merger with a company on a different platform forces a collaboration change, sign-in for every connected app comes along for the ride.

We treat identity and device management as one of seven distinct layers in an architecture, alongside collaboration, security monitoring, backup and portability, business applications, physical infrastructure, and connectivity. Designing identity as its own layer means writing down why you chose your provider, which applications depend on it, and what it would take to move. That last part is your back-out path.

Choosing an identity provider on purpose

There are several credible options, each with different strengths.

  • Microsoft Entra ID
    • Often a good fit when: You run Microsoft 365, Windows devices, and Intune, and want tight integration
    • Questions to ask: Which features need P1 or P2 licensing? How much of your policy depends on Microsoft-only signals?
  • Okta
    • Often a good fit when: You run a mix of SaaS apps and collaboration platforms and want identity independent of any one of them
    • Questions to ask: How will device signals and Microsoft 365 sign-in be handled? What does it cost at your scale?
  • Google Cloud Identity
    • Often a good fit when: You run Google Workspace or Chromebooks and want Google-managed identity
    • Questions to ask: How well does it handle your Windows devices and on-premises applications?
  • RSA
    • Often a good fit when: You already rely on RSA authentication or need strong authentication for on-premises and regulated systems
    • Questions to ask: How will it work alongside your primary directory and single sign-on provider?

For many Microsoft 365 organizations, Entra ID is the right answer. The goal is to make that choice with eyes open, document it, and know how you would change course if you had to. If you still run an on-premises directory and are weighing what comes next, replacing Active Directory walks through the options.

Questions to bring to your next planning meeting

  • Which applications still have their own passwords outside your identity provider?
  • Which accounts can approve payments, change access, or administer systems, and do they use phishing-resistant MFA today?
  • Can a personal, unmanaged device reach sensitive data, and did someone decide that on purpose?
  • How long does it take to fully remove a departing employee's access, and who verifies it?
  • If you changed email and collaboration vendors in three years, what would happen to sign-in for everything else?

Start with the foundation

Identity sits underneath every other security control you own. Getting it right is mostly disciplined design: one authoritative provider, strong authentication, device-aware policy, clean lifecycles, and a documented reason for every major choice.

At CTP, we design identity as the foundation of the architecture and keep it separate enough that your next collaboration decision stays yours to make. To see how that works in practice, read How We Design, or explore security and access solutions built around that approach.

Sources and further reading

Ask about something important for you.

Curious about our take on an industry or technology topic? Let us know.

Contact Us