Blog
Conditional Access policies across MSP tenants
Standardize Conditional Access policies across 20 to 100 client tenants. Learn baseline licensing, report-only testing, and drift detection.
Nerdio is Microsoft’s 2026 Americas ISV Partner of the Year Get the story
Blog
Standardize Conditional Access policies across 20 to 100 client tenants. Learn baseline licensing, report-only testing, and drift detection.
Table of Contents
Conditional Access policies live inside each client's Entra ID tenant, so managed service providers (MSPs) need a cross-tenant process to copy and sync them at scale.
Six weeks ago, a technician disabled a client's "Require MFA for all users" policy to troubleshoot a sign-in failure and never re-enabled it. Nothing alerted. Now the client's cyber insurer wants written proof that MFA is enforced on every account, while the other 47 tenants run whatever baseline version was current at onboarding.
This guide is for MSP owners, practice leads, and service delivery leaders managing 20 to 100 Microsoft 365 client tenants.
Conditional Access is Microsoft's Zero Trust policy engine. According to Microsoft, it combines signals to make and enforce access decisions. Each policy works as an if-then statement. If a user wants to access Microsoft 365, they must perform multifactor authentication.
Policies apply after first-factor authentication and then either allow access, sometimes with session controls, or block it.
Every client should start with MFA and legacy-authentication controls, subject to licensing. Microsoft's Digital Defense Report 2026 attributes over 99% of observed cloud initial access efforts to password spray. The 2025 edition names phishing-resistant MFA as the most effective control against compromised credentials and found MFA blocks access in over 99% of cases even when attackers hold a valid username and password.
Microsoft's "Secure foundation" template group, recommended as the base for all organizations, includes Require MFA for all users and Block legacy authentication.
For an MSP security baseline, these findings put credential controls ahead of more specialized policies. Each client's licensing then sets the ceiling on which of those controls you can actually deploy.
Conditional Access requires Entra ID P1. Microsoft 365 Business Premium and Microsoft 365 E3 include P1; Business Basic and Business Standard do not, so those tenants fall back to Security Defaults, which offer basic protection with no customization.
Risk-based policies (sign-in risk and user risk) require Entra ID P2, and Business Premium does not include P2. For the MSP, that licensing split determines which policy package can be deployed. It also determines whether risk controls belong in the baseline or in a P2-only variant.
Business Premium stayed at $22.00 per user per month in Microsoft's July 1, 2026 update. Microsoft's Entra pricing page listed Entra ID P1 standalone at $7.00 as of October 2026. The standalone P1 price gives Basic and Standard customers a separate upgrade path without changing their Microsoft 365 plan.
Creating any Conditional Access policy prevents enabling Security Defaults, and Microsoft advises turning on Conditional Access immediately after disabling them. If a client's P1 licenses lapse, existing policies remain, but they cannot be updated, which freezes that tenant out of your next baseline revision.
Nerdio Manager for MSP's Tenant Assessment wizard (Public Preview as of v7.0) scans a customer or prospect tenant's existing Conditional Access policies and Microsoft Intune profiles and compares them with your baselines.
The safest baselines start from Microsoft's templates and stay in report-only until real sign-ins prove them. Below we'll cover how to keep the baseline small enough to manage across every tenant, how to stage and test it against real traffic before enforcement, and how to protect break-glass accounts and track every exclusion.
Microsoft's deployment plan caps a tenant at 240 policies in any state and recommends grouping apps with identical requirements into a single policy. A tight baseline leaves room for client-specific additions without hitting that limit. Microsoft's architecture guidance adds a naming convention, which matters across dozens of tenants because consistent names make policies easier to audit, compare, and troubleshoot during drift reviews.
A workable baseline derived from Microsoft's templates and deployment plan includes the following policies:
This set addresses credential attacks, risky sign-ins, unsupported devices, and persistent sessions while keeping the baseline well under the 240-policy limit. Treat it as a derived MSP baseline rather than a Microsoft-published package, and confirm each tenant's licensing and dependencies before deployment.
In inherited tenants, you may find classic policies, the pre-Entra Azure AD Conditional Access policies Microsoft retired. They stopped enforcing after July 10, 2024, so anything still in that list is dead weight and should be removed before the new baseline goes in.
Template policies are created in report-only by default, and Microsoft recommends at least one week there before switching on. The What If tool is useful but doesn't test service dependencies (a Teams test ignores Exchange Online policies), and Microsoft's own plan says a simulation "doesn't replace an actual test run in a properly configured development environment."
Pair every change with a pilot group and a named rollback owner. The pilot group catches policy conflicts and user-impact issues on a small, known population before the change touches the whole tenant, and the named owner makes sure someone is accountable for reverting if sign-ins start failing.
To support that rollback, export each template as JSON before enforcement and re-import with Upload policy file, which gives you a per-tenant rollback artifact you can restore in minutes.
For Entra ID B2C tenants, Microsoft recommends two or more break-glass (emergency access) accounts. It also recommends excluding emergency accounts from every policy that blocks or restricts sign-in; a break-glass account subject to an MFA or compliant-device policy may be unusable during the emergency it exists for.
Microsoft also recommends an Azure Monitor or Sentinel alert on every sign-in, using a log search rule keyed to each account's Object ID. Once enforced, mandatory admin-portal MFA applies to break-glass accounts regardless of Conditional Access exclusions, so Microsoft recommends passkey (FIDO2) or certificate-based authentication for them.
Each exclusion should have an owner and a review date, with Entra access reviews providing evidence for auditors that exceptions get revisited.
The harder job is multi-tenant policy management to get that baseline into every tenant and keep it there.
No single Microsoft tool deploys and governs Conditional Access across every tenant on its own. Each option covers part of the job, and the right mix depends on how much per-tenant deployment work and ongoing drift review your team is prepared to own. Evaluate each multi-tenant management option by how it deploys a common policy across tenants, translates tenant-specific references, handles consent, monitors drift, and preserves intentional per-tenant differences.
Microsoft 365 Lighthouse is available to CSP-enrolled partners at no additional cost, and its default baseline includes "Require MFA for admins" with a sub-task to create a Conditional Access policy. Deployment runs one tenant at a time through Tenants > select tenant > Deployment Plan > select task > Deploy, which gives the MSP tenant-level control over each deployment. Tenants over 2,500 licensed users get only limited experiences, and Microsoft notes that detecting an enabled MFA policy doesn't guarantee all targeted users use MFA.
When Lighthouse falls short, Microsoft Graph and PowerShell can create policies with New-MgIdentityConditionalAccessPolicy under the Policy.ReadWrite.ConditionalAccess permission. Policies reference groups and named locations by tenant-specific GUID, and Microsoft-provided scripts do not remap those references. MSPs have to translate them before using an exported policy in another tenant. Under GDAP, application-only permissions also need manual tenant-wide admin consent in each customer tenant. Remapping and consent become per-tenant deployment steps every time.
Two more Microsoft options sit outside that per-tenant workflow:
Nerdio Manager handles the remapping problem by using variables in the JSON policy definition. One policy carries per-tenant differences such as named locations without being rewritten for each customer.
Even a baseline that lands cleanly in every tenant will drift over time. Technicians, Microsoft, and your own exceptions all introduce changes that pull individual tenants away from the standard. The three sub-sections below cover where drift comes from, how to detect it per tenant, and why investigation at scale strains native tooling.
Two common vectors of drift inside a single tenant are:
Drift becomes MSP liability when a breach traces to a deviation from the agreed standard. The multi-tenant version is where a policy gets updated for one client and never applied to the rest.
Microsoft itself is also a drift source. Microsoft-managed policies arrive in report-only and switch on no less than 30 days later unless an admin sets them Off. From June 15, 2026, policies targeting "All resources" with resource exclusions began evaluating sign-ins that request only baseline scopes. The rollout ran progressively over several weeks per tenant, and Microsoft enabled it automatically wherever baseline-scopes settings were unchanged. Between July 6 and July 13, 2026, policies scoped to Register security information started evaluating during Windows Hello for Business and macOS Platform SSO credential registration.
For an MSP managing 50 tenants, that progressive June rollout could reach different tenants at different times, which produces the same end-state as a technician making one-off changes.
Secure Score gives you a per-tenant posture signal to catch drift as it happens. Microsoft Secure Score awards partial credit by percentage, so 50 of 100 users protected by MFA earns 5 points. Microsoft recalculates Identity Secure Score every 24 hours, and an action marked "Risk accepted" earns no points. A score drop in a single tenant is a drift flag worth investigating.
Detecting a Secure Score drop is the easy part. Tracing that drop back to the specific policy change, exclusion, or Microsoft-managed rollout that caused it is where native tooling runs short. Lighthouse offers a comparative view across managed tenants, but tracing a drop to a specific policy change usually means opening that tenant. Across 50 tenants, each drop adds a per-tenant manual step inside a flat monthly fee, affecting your MSP operational efficiency.
Nerdio Manager closes the per-tenant deployment and drift-tracing gaps left by Lighthouse, Graph, and Secure Score. It is a Microsoft-focused management and automation layer that extends Microsoft's cross-tenant management capabilities. Policies still run in each client's Entra ID, and Microsoft remains the enforcement engine.
MSPs push a master template of Conditional Access policies from a single console under Account level > Policies > Entra > Conditional Access, and manage per-customer variants in the same place as part of Microsoft 365 management. A baseline revision gets made once in the master template and applied across customers, rather than rebuilt separately for each one.
For Intune Policy Baselines, new or updated policies are reflected across customer tenants "within the hour". The temporary policy created when validating a Conditional Access policy is set to Report-only to prevent self-lockout.
Drift controls run on the same deployment workflow. Solution Baselines package and deploy standardized configurations across customer environments and detect policy drift against the assigned baseline. Account-level variables preserve intentional per-customer differences, which separates approved variations from unmanaged changes. Feature rollback supports recovery when a change needs to be reverted.
For insurer reporting, Nerdio Manager centralizes policy management and audit logging across customer environments, creating records of policy administration that answer the "prove MFA is enforced" question without opening each tenant.
Kilpatrick IT reported a 200% ROI after using Nerdio Manager to migrate 40 tenants 66% faster, the onboarding speed that matters when every new client has to land on the same baseline from day one.
To close the six-week MFA blind spot and answer the insurer's request for written proof, inventory every tenant against one baseline. That inventory identifies the disabled policy and which of the other 47 tenants are not on the current baseline version while creating initial policy-state evidence.
Because policy state does not prove that every targeted account uses MFA, confirm the insurer's required evidence format and include evidence for assignments, exclusions, and sign-in records.
From there, four steps move the whole estate onto the same baseline without disrupting end users.
To see how Nerdio Manager handles the master-template deployment, drift detection, and audit evidence described above across every client tenant, get a demo or walk through the console in the self-guided demo.
Yes. Each policy exists only in the tenant that owns it, so standardization means deploying and monitoring a template in every tenant. Microsoft 365 Lighthouse deploys one tenant at a time, and Graph exports require remapping for tenant-specific GUIDs. Nerdio Manager applies per-tenant variables and flags drift against the assigned policy.
A client tenant needs Entra ID P1, which Microsoft 365 Business Premium and Microsoft 365 E3 include. Business Basic and Business Standard tenants rely on Security Defaults instead, and risk-based sign-in and user policies require Entra ID P2. For Microsoft-managed policies specifically, Microsoft's managed-policies page lists "Microsoft Entra ID P2 or Microsoft 365 Business Premium" as its prerequisite. Microsoft does not reconcile this prerequisite with its P1 guidance for Conditional Access.
Microsoft's deployment plan sets the limit at 240 policies in any state. Grouping apps that share requirements into one policy, and naming policies consistently, keeps the set manageable across dozens of tenants.
No. They cover a subset of recommended controls and switch on at least 30 days after introduction unless set to Off. Admins can only exclude users, such as break-glass accounts, or change state, so custom edits require a duplicate policy.
Learn more about Nerdio Manager