Skip to main content

Nerdio is Microsoft’s 2026 Americas ISV Partner of the Year Get the story

Blog

Microsoft Intune management: how to run endpoint operations at scale

Learn how Windows 365, Intune, and Azure Virtual Desktop management scales patching, policy, app delivery, automation, and reporting across your fleet.

Take a 6,000-endpoint Windows fleet on the Wednesday after Patch Tuesday. Only 71% are on the latest quality update, and nobody can explain the rest. A BitLocker profile edit just left two devices stuck on error 0x87d101f8. Enterprise endpoint management with Microsoft Intune gets hard fast when you need to trace a policy change or land an app before the next check-in.

Running Intune at this scale takes one operating model that coordinates patching, policy governance, application delivery, Microsoft Graph automation, and reporting across physical endpoints and Windows Cloud (the umbrella for Windows 365 and Azure Virtual Desktop).

This guide shows end-user computing leaders, Intune owners, and IT operations teams how to build that operating model.

What Intune management covers at enterprise scale

At enterprise scale, Microsoft Intune uses Microsoft's cloud service for unified endpoint management to enroll devices and deliver applications while applying configuration and compliance policy. Entra ID supplies identity signals for access decisions.

In the 6,000-endpoint scenario, one operating model coordinates those controls and workloads across the fleet. Mobile device management (MDM) controls the whole device while mobile application management (MAM) controls corporate data inside apps on hardware you may not own. Compliance results flow to Entra ID, where Conditional Access decides whether a device reaches your resources.

Many enterprises use Windows 365 and Azure Virtual Desktop together and manage them alongside physical PCs. Windows Cloud can operate from a single Intune tenant, including supported Azure Virtual Desktop session hosts. This shared tenant keeps device policy and application operations together across endpoint types. Windows 365 Enterprise Cloud PCs enroll automatically during provisioning, with no user credentials involved. Pooled multi-session hosts can enroll too, with a narrower policy set.

Settings catalog and compliance policies apply to those hosts, but Microsoft does not currently support Windows update rings, and Intune reports unsupported device configuration templates as Not applicable. A management platform can schedule quality updates for those pooled multi-session hosts.

Co-management and tenant attach for fleets still on Configuration Manager

Enterprises running Configuration Manager can co-manage the same Windows devices with Intune, so workloads move independently rather than in a cutover. Tenant attach surfaces those devices in the Intune admin center with a unified device timeline, where administrators can run CMPivot and remote actions.

For the illustrative 6,000-endpoint fleet with on-premises management still in place, this provides a Configuration Manager migration path.

What the July 2026 licensing bundling means for your tenant

The July 1, 2026 bundling redefined what "native" means for your tenant. Microsoft bundled Intune Suite capabilities into Microsoft 365 E3 and E5, with rollout complete by August 1, 2026.

License

Intune capabilities included (from July 1, 2026)

US list price, per user/month, annual

Intune Suite (add-on)

Plan 2, Endpoint Privilege Management, Cloud PKI, Advanced Analytics, Remote Help, Enterprise App Management

$10.00

Microsoft 365 E3

Plan 1, Plan 2, Remote Help, Advanced Analytics

$39.00 (was $36.00)

Microsoft 365 E5

E3 capabilities, Endpoint Privilege Management, Cloud PKI, Enterprise App Management

$60.00 (was $57.00)

 

Prices are Microsoft's September 2026 US list figures; Enterprise Agreement and CSP pricing differ.

For E3 and E5 tenants, the budget question has shifted from whether to buy Intune Suite to how to run what just arrived, and patching is the first workload that exposes the answer.

 

Patch at scale with Autopatch, hotpatch, and coverage for the gaps

Getting to a defensible compliance number across physical PCs, Cloud PCs, and session hosts takes three moving parts working together. Autopatch sets the baseline for eligible devices, hotpatch removes restarts when the prerequisites line up, and a management platform handles the multi-session hosts Autopatch will not touch.

Autopatch baseline and prerequisites for E3 and E5 tenants

For the E3 and E5 tenants in that table, Windows Autopatch is already licensed, and it targets keeping at least 95% of eligible devices on the latest quality update.

Devices must be corporate-owned and Entra joined or hybrid joined. Entra registered devices are not supported. Update ring policy still sets deferral periods and deadlines, and critical fixes can override those deferrals.

Hotpatch cuts restarts when every prerequisite lines up

Hotpatch changes the restart math. Windows Autopatch has enabled it by default for eligible devices since May 2026. Microsoft's Windows update documentation states that "Applying security fixes without waiting for a restart can get organizations to 90% compliance in half the time."

A baseline cumulative update with a restart lands in the first month of each quarter, then restart-free hotpatches in months two and three. Every prerequisite has to hold at once:

Devices that are ineligible keep receiving standard monthly updates. Windows 365 Cloud PCs get full Autopatch parity with physical devices, hotpatch included.

Cover the gaps Autopatch leaves on multi-session hosts

Azure Virtual Desktop splits by session type. Autopatch supports single-session hosts but not multi-session hosts. Quality updates on pooled multi-session hosts therefore run through settings catalog policy or a separate process.

Nerdio Manager for Enterprise covers that lane. Scripted Actions include built-in Windows patching tied to Patch Tuesday schedules, so hosts Autopatch will not touch stay on a schedule. Howell Technology, a UK managed service provider (MSP) running Nerdio Manager, reported monthly security patching dropping from 8 hours to 2 hours per customer.

Patching succeeds or fails on a schedule you can see in a report. Policy failures can be easy to miss without active review, which is why policy needs its own design discipline.

Policy conflicts and drift require active review

Patch reports tell you which device missed. A policy conflict can go unnoticed unless you review the generated errors. When two policies set the same setting to different values, Intune generates an error and applies neither. The manual fix is to find the overlapping profile and set the redundant setting to Not Configured. The BitLocker error 0x87d101f8 from the opening scenario is exactly this pattern, two profiles configuring the same encryption setting.

Several Intune defaults require attention at scale. When devices have no compliance policy assigned, Intune marks them Compliant by default, and Microsoft recommends flipping that to Not Compliant before Conditional Access starts trusting the wrong signal. The second gap is disaster recovery. Policy backup and restore require Microsoft's documented Graph API scripting, and a complete tenant backup goes beyond what the portal can produce.

The community IntuneBackupAndRestore module fills part of that gap, but it can only restore a policy that no longer exists in the portal. Restoring an edit to a policy that still exists means running your own Graph scripts or using a management platform that keeps versioned backups.

Administrative access, audit trails, and elevation

Access design is the other half of policy governance. Administrators assign Intune roles to groups, and scope tags limit which objects an admin can see. Microsoft caps tags at 100 per role and 100 per object. Intune RBAC does not constrain Entra roles, so an Intune Service Administrator has full access regardless of tags.

Microsoft's March 14, 2026 guidance on securing Intune recommends removing broad role assignments and scoping administration by business unit or region. It also recommends Entra PIM for time-bound elevation. Audit logs keep every change for two years and no admin can delete them. Longer retention means routing them to Azure Monitor.

Endpoint Privilege Management arrived in Microsoft 365 E5 with the July 1, 2026 bundling, so E5 tenants already own it. Endpoint Privilege Management keeps users as standard users by default. It elevates specific applications or tasks per rule or on demand and records each elevation. That supports removing standing local admin rights across the 6,000 endpoints in this scenario while allowing configured applications and tasks to be elevated.

Policy defines what a device should look like. Applications are where the wait between "assigned" and "installed" turns into a ticket.

Cut app delivery time below the native check-in cycle

Native Intune delivery runs on device check-in and group processing schedules, which sets a floor on how fast an app can land. Two mechanics set that floor (the check-in cadence itself and the group targeting behind it), and a management platform on top of Graph can push delivery below it.

Check-in cadence sets the floor on how fast an app lands

A Windows device checks in every 3 minutes for 15 minutes, then every 15 minutes for 2 hours, then roughly every 8 hours, plus one maintenance sync every 6.5 hours. After 24 hours without check-in a device cannot receive policies. The Intune Management Extension that installs Win32 apps checks hourly or on restart, and a single app can reach 30 GB.

Group targeting and multi-session hosts add their own lag

Even when a device is checking in on schedule, the targeting layer decides when it actually sees a new assignment. Dynamic Entra ID group changes usually process within a few hours, and Microsoft documents processing that can exceed 24 hours in large tenants. The combined ceiling of 15,000 dynamic groups and dynamic administrative units per tenant cannot be raised, which forces design discipline once the estate grows. Assignment filters evaluate at check-in and sidestep the group wait, though they still depend on the check-in cadence to schedule anything.

Multi-session hosts add a separate set of constraints on top of that targeting layer. Apps must install in device context with Required or Uninstall intent, and Company Portal cannot install apps on multi-session hosts at all. Any app strategy that leans on user-driven installs has to route around those hosts.

Push apps below the check-in floor with Graph orchestration

Native delivery operates on assignment and check-in timing, which means an app may not land immediately. Nerdio Manager's Unified Application Management deploys the same app to a Cloud PC in roughly 30 seconds because it does not wait for the next poll.

On Intune-managed physical devices, deployment runs per the Intune agent's scripted action assessment schedule. You can script Intune policy, patching, and application-delivery operations against Microsoft Graph, and Graph has ceilings of its own.

How to extend Intune automation with orchestration

Every Intune admin center action has a Graph API equivalent, which is what makes fleet-wide automation possible in the first place. Four practices extend Intune automation into a reliable operating model on top of Graph:

  • Design for Graph throttling:
    The Intune device service allows 200 write requests per 20 seconds across all apps and 100 per app. Exceed that and Graph returns HTTP 429 with a Retry-After header, and the SDKs do not retry throttled requests inside a JSON batch. A bulk change becomes a queueing problem, and your orchestration needs retry logic rather than a single script call.
  • Route automation through Multi Admin Approval:
    Multi Admin Approval now covers automation and requires the same approval when a service principal modifies a protected resource as when a human admin does. Any orchestration that touches protected resources has to fit into the approval flow instead of around it.
  • Wrap remediations and exports in a scheduler:
    Remediations run detection and remediation script pairs on a schedule, though the on-demand "Run remediation" action targets a single device. The native portal can export configuration settings to CSV, which works for one-off operations. Fleet-wide recovery and repeatable execution benefit from orchestration around Graph.
  • Add AI-assisted review at the operator layer:
    Copilot in Intune works the operator side of the same data. It summarizes what a policy or setting does in plain language and flags conflicts while you review a policy, and AI agents handle Multi Admin Approval triage and vulnerability prioritization.

A management platform that sits on Graph can add scheduling and retry orchestration alongside your scripts. Nerdio Manager works through the same Graph permissions your scripts already request.

How Nerdio Manager extends Intune across the fleet

Nerdio Manager reads and writes Intune objects through Graph. Intune stays the system of record and the policy store. The platform deploys into your Azure tenant and operates through permissions such as DeviceManagementConfiguration.ReadWrite.All and CloudPC.ReadWrite.All. Monitoring and reporting require no proprietary Nerdio monitoring agent on session hosts.

On the Windows 365 side, Nerdio Manager's strongest contribution is extended Intune management for each Cloud PC, covering Intune integration, reporting, policy operations, and application management. Those capabilities live in the same console that manages Azure Virtual Desktop, which matches how many enterprise customers actually run their Windows Cloud estate across both products.

Policy backup, restore, and conflict resolution

Every edit to an Intune policy or profile triggers an automatic backup, including edits made in the native Intune console. You can restore a policy by itself or with its assignment. You can also restore assignments only and compare two versions side by side with the differences highlighted. Policy Studio replaces raw JSON editing with a building-blocks view.

Policy Studio maps policies to the devices and groups they reach. It also flags settings where two assigned policies collide. A policy approval workflow requires a designated approver to review every change request before it reaches Intune. Syneos Health reported 70% faster virtual desktop configuration changes after adopting Nerdio Manager.

In TechTarget's economic validation, one customer told them: "We work differently now that we have Nerdio. The relationship between IT and business is more of a partnership because we now have the time to better understand their needs and find solutions to help them reach their goals."

Application delivery and Cloud PC operations

Unified Application Management links WinGet and Intune repositories into one catalog and pushes apps by policy to every Windows device you manage, Cloud PCs and multi-session hosts included. Deployment to Intune-managed devices uses policy-based assignments and covers MSI and EXE packages in device context.

For Windows 365, the same console brings those capabilities into the wider Windows Cloud view and adds right-sizing recommendations.

Fleet reporting and monitoring

You build Intune Insights dashboards from widgets configured for compliance and patch status across the whole fleet, so Intune reporting matches how your team splits the estate.

Nerdio's Real-Time Insights has live performance data for physical Intune-enrolled endpoints, alongside the existing Cloud PC and session host views. You can configure the polling interval down to one minute from the five-minute default. 

LexisNexis, which serves customers in 160 countries, reported a 75% decrease in support needs after deploying Nerdio Manager across its organization.

Nerdio Manager schedules and deploys workflows that would otherwise require Graph scripts. It also handles backup and monitoring. You can track patch compliance and device health to judge whether that orchestration is paying off, then check the effect on admin effort.

What to measure once the automation is running

That Wednesday after Patch Tuesday needs a scorecard, and the 71% compliance figure from the opening scenario is only the first line on it. Four measures track whether the operating model is holding:

  • Patch compliance against Autopatch's 95% target, split into hotpatch-eligible and ineligible devices so a restart backlog does not hide inside an aggregate.
  • Device compliance rate, which means something only after the unassigned-device default is set to Not Compliant.
  • Intune's Endpoint Analytics score, a 0–100 composite of startup performance, application reliability, and Work from Anywhere. It sits at an industry median of 50 and needs at least five devices to register.
  • Help desk and per-device admin effort, tracked against industry benchmarks such as first contact resolution and cost per contact. Routine per-device management effort varies depending on automation maturity, and across the illustrative 6,000-device fleet, small per-device inefficiencies compound into substantial admin time. A BitLocker conflict or a three-hour app wait increases that burden.

If your Wednesday looks like the one at the top of this article, you can put orchestration over the Intune you already own, with Cloud PCs and multi-session hosts in scope from day one. Get a demo to see how Nerdio Manager works across your Windows 365, Intune, and Azure Virtual Desktop environment, or try it free in your Azure tenant.

Frequently asked questions about Intune management

Ready to get started?