Skip to main content
Jeremy Ingram
January 7, 2026
Revised: March 23, 2026

Compelling Factors

At the time of this blog, it is December 2025: My phone does not have an AI agent to carefully curate Ferroque’s weekly reports and Visio diagrams into a beautiful annual summary, but if it did, surely one of this year’s project themes would be customers taking advantage of their M365 licensing and piloting their workloads on Azure Virtual Desktop (AVD) and Windows 365 (W365).

When you think about it, this is highly significant for several reasons. Migrating desktops to the cloud is nothing new, but in 2025, there was a perfect storm of customer justifications for migrating desktop workloads into Azure desktops:

  • Licensing and Commercial Pressure
    • Broadcom acquisition impact on VMWare licensing.
    • Increased licensing costs and packaging complexity.
    • Desire to reduce overlapping vendor spend.
  • Infrastructure Refresh Avoidance
    • Avoiding capital investment in new hypervisor hardware.
    • GPU scarcity and cost inflation.
    • Data center consolidation or exit strategies.
  • Security, Identity, and Zero Trust Alignment
    • Tighter integration with Entra IS and modern authentication.
    • Reduced attack surface on endpoints.
    • Simpler alignment with Zero Trust principles.
  • Workforce and Usage Pattern Changes
    • Growth of remote, hybrid, and distributed workforces.
    • Increase in contractors, partners, and third parties.
    • Shift-based and frontline work models.
  • Disaster Recovery and Business Continuity
    • Desire for simpler desktop DR strategies.
    • Regional resilience and geographic flexibility.
    • Operational resilience without full platform duplication.
  • Cost Model Preference
    • Shift from CapEx to OpEx.
    • Ability to right-size desktops over time.
    • Improved cost transparency.
  • Strategic Alignment with Microsoft Roadmap
    • Confidence in Microsoft’s long-term investment in cloud desktops.
    • Simplification of vendor relationships.
    • Alignment with broader Microsoft security and AI initiatives.

Designing with Intent

Regardless of the drivers, most conversations about AVD and W365 begin with deceptively simple questions, typically along the lines of:

  • “Should I use AVD or W365?”
  • “When should I use Windows 365?”
  • “How is Windows 365 different from AVD?”

Default design decisions are used when organizations make a cloud desktop decision before adequately defining target use cases, and those default design decisions can inadvertently become production architecture if/when redesign and redeployment are deemed too expensive or operationally difficult to complete. Over time, these default decisions often lead to cost inefficiencies, operational friction, scaling challenges, and user dissatisfaction; not necessarily because the platform failed, but because the original design intent was never clearly articulated and considered before build/deployment activities occurred.

AVD and W365 are both strategic Microsoft offerings, but they are not interchangeable. Each was built to address different scenarios under different constraints. Treating these products as equivalents or forcing one to behave like the other typically results in a compromised solution.

“Designing with intent” for AVD and W365 is no different from designing for other EUC technologies, wherein fulfilling end-user requirements and delivering a good user experience are of the utmost importance. In any EUC design effort, the user personas, endpoints, and peripherals drive the requirements for the overarching design. It has been our experience that when the user personas are defined up front, the best-fit platform(s) begin to come into sharper focus.

AVD and W365: Different Tools, Different Intent

AVD and W365 are often discussed as alternatives for delivering Windows desktops on Azure. While this is technically true, it is architecturally misleading. The two services are designed around fundamentally different assumptions regarding control, responsibility, and operational ownership.

Understanding those assumptions is critical because most design failures occur when an environment is built under the wrong model.

AVD Requires More Design Decisions

AVD is designed for scenarios in which organizations need to make and retain explicit architectural and operational decisions. Compared to W365, AVD requires more upfront and ongoing design choices, but in return provides greater flexibility in how VDI environments are built, scaled, and optimized.

In practice, choosing AVD means architects must deliberately design elements such as:

  • Host Pool Architecture
    • Pooled/persistent versus personal/non-persistent desktops.
    • Multi-session vs single-session workloads.
    • Host sizing and user density assumptions.
  • Scaling behavior
    • Power management schedules.
    • Scaling workloads to peak and off-hours usage.
    • Considerations for spikes in resource demand.
  • Image lifecycle
    • VDI image update cadence.
    • Validation and rollout processes.
    • Rollback and recovery strategies.
  • Networking and security integration
    • Virtual network design, routing, and DNS.
    • Integration with existing security controls.
    • Connectivity to on-premises private resources.

The above decision points heavily influence consumption costs, performance, and user experience. For organizations that have hosted carefully curated workloads on EUC platforms such as Citrix DaaS and are accustomed to the administration commitment of a similar application and desktop delivery platform, such workloads are often good candidates for migrating to AVD. For less complex workloads, a W365 solution may be a more appropriate delivery platform.

W365 Requires Fewer Design Decisions

W365 takes a deliberately different approach. It is designed to reduce the number of architectural decisions customers must make, favoring a pre-defined service model that prioritizes consistency, predictability, and operational simplicity.

With W365, architects do not need to design elements such as:

  • Host pools or scaling logic.
  • VM start/stop schedules.
  • Session density or concurrency behavior.
  • Infrastructure placement or capacity planning.

Instead, W365 provides a service in which:

  • A cloud PC is provisioned per user (or per Frontline entitlement).
  • Microsoft manages capacity and availability.
  • Desktops are persistent and predictable by default.
  • Lifecycle and policy are managed via Intune.

This approach aligns well with organizations that want desktops to behave more like managed endpoints rather than a platform that must be continuously tuned. The absence of infrastructure-level controls is intentional and removes an entire class of operational responsibility from IT teams.

Mismatched Platforms and Use Cases

A frequent pattern we see in the field is deploying AVD for use cases better suited for W365, such as:

  • Always-on, dedicated desktops per user.
  • Minimal operational overhead expectations.
  • Rapid provisioning without platform planning.

While AVD can be adapted to approximate these outcomes, doing so often results in over-provisioned environments, unnecessary complexity, and less predictable cost behavior, the inverse mistake also occurs, of attempting to use W365 for scenarios that require infrastructure-level control, only to discover that those controls are intentionally unavailable.

In both cases, the issue is not the platform itself but a mismatch between the use case and the design intent. Here is the most important distinction to internalize:

  • Choose AVD when your requirements demand architectural control and customization.
  • Choose Windows 365 when your requirements favor simplicity and reduced design responsibility.
  • Use both when different user populations clearly require different outcomes.

These choices determine how environments scale, how incidents are resolved, how changes are introduced, and the long-term sustainability of the platform. Misunderstanding this distinction at the onset almost always leads to re-architecture down the road.

The following table provides a side-by-side overview of AVD and W365 use cases.

Azure Virtual DesktopWindows 365
AVD is best suited for scenarios wherein control is considered a key feature. Common use cases that align well with AVD include:
• Multi-session desktops to optimize costs per user.
• Customer networking and security architectures, such as hub-and-spoke designs or private connectivity.
• Specialized workloads, including GPU-enabled desktops or latency-sensitive applications.
• Advanced cost optimization.

AVD also aligns well with organizations that already operate complex enterprise infrastructure and are prepared to own image lifecycle, scaling logic, and cost governance.

Many customers are already eligible to use AVD from a licensing perspective, but the eligibility does not eliminate the need to design for Azure consumption. Compute, storage, and networking costs remain real architectural considerations and require strategic and intentional design.
W365 Enterprise is designed for organizations that want a Cloud PC operating model, wherein desktops behave like a managed service. W365 is a strong fit when:
• Users need persistent, predictable virtual desktops.
• Operational simplicity is prioritized over infrastructure optimization.
• Endpoint management is standardized around Intune and Entra ID.

W365 Enterprise provides a dedicated Cloud PC per user, available 24×7 and tightly integrated into Microsoft’s endpoint and security ecosystem.

W365 Frontline brings this model to shift-based and non-concurrent usage patterns, wherein users need access during working hours but do not require a dedicated always-on Cloud PC. Frontline is particularly well-suited for:
• Call centers and service desks.
• Healthcare, retail, and manufacturing environments.
• Seasonal, part-time, or rotating staff.
• Shared working scenarios.

By aligning Microsoft licensing to real-world usage patterns, Frontline helps organizations avoid over-provisioning desktops for users who only need intermittent access.

When the Requirement is Apps, not Desktops

In some scenarios, instead of a full VDI desktop, the requirement is access to a small set of applications. W365 Cloud Apps (currently in preview) reflects Microsoft’s recognition that app-first delivery remains a strategic application delivery model, especially for frontline and task-based workflows.

Cloud Apps further reinforces Microsoft’s move toward right-sized access rather than one-size-fits-all desktops, and should be evaluated strategically and carefully aligned with usage and delivery scenarios.

Of course, publishing apps from Azure-hosted desktops is nothing new; this functionality has been available since AVD was first introduced. Across AVD and W365 platforms, there are important design considerations to keep in mind. Single-session/multi-session limits still apply to applications that are published to users.

  • User and application density are extremely limited, with applications published on W365 instances at a 1:1 ratio.
  • Increased user density is available with applications published on AVD session hosts, which can scale with changes in user demand throughout the day.

Operating Models Day-2 Reality

One of the most consistent predictors of success or failure in Azure-hosted desktop environments is not the platform choice itself, but whether the operating model matches the platform’s design intent.

AVD and W365 both deliver Windows desktops, but they assume very different day-2 responsibilities. When those responsibilities are not clearly understood or are underestimated, environments tend to degrade gradually rather than fail catastrophically. By the time issues surface, they are often expensive to correct.

A key differentiator is that AVD assumes the organization is prepared to operate a platform, not just consume a service. Beyond initial deployment, IT teams must continuously manage and coordinate several moving parts. Alternatively, W365 shifts the operational center of gravity away from infrastructure and toward endpoint, identity, and policy management.

The following table provides a side-by-side overview of AVD and W365 Day-2 operations.

Azure Virtual DesktopWindows 365
Image lifecycle and change management
AVD environments require an explicit image management process:
• Image updates must be planned, tested, and rolled out deliberately.
• Application changes must be validated against pooled or multi-session hosts.
• Rollback procedures must exist for failed updates.

When image ownership is unclear, environment drift quickly leads to inconsistent user experiences, unplanned outages, and/or prolonged remediation.
Endpoint lifecycle and policy
Windows 365 Cloud PCs behave much like physical endpoints:
• Configuration and compliance are managed and enforced via Intune.
• Security posture is maintained through identity and policy.
• Updates follow standardized endpoint management workflows.

This naturally aligns with organizations that already operate mature endpoint management practices.
Capacity planning and scaling behavior
AVD does not automatically “do the right thing” without guidance:
• Host pools must be sized correctly for concurrency assumptions.
• Autoscaling logic must be configured and maintained.
• Peak usage patterns must be revisited as user behavior changes.

If scaling logic is static or poorly tuned, organizations often experience either persistent over-provisioning or recurring performance complaints during peak periods.
User experience consistency
Since capacity, availability, and persistence are handled by Microsoft:
• Desktops behave consistently across users, if there is consistency in VM specs across hosts.
• Performance variability is reduced.
• Operational surprises related to scaling are minimized.

This predictability is often more valuable than raw flexibility, particularly for executive, mobile, or frontline users.
Cost governance
Because AVD is consumption-based, cost control must be an ongoing operational discipline, not a one-time design task:
• Idle resources must be identified and addressed.
• Scaling policies must align with real usage, not assumptions.
• Storage and performance tiers must be periodically reviewed.

Without ongoing cost ownership, AVD environments tend to become more expensive over time; not because the platform is inefficient, but because optimization levers are no longer actively managed.
Reduced platform ownership
W365 intentionally removes entire classes of operational responsibility, for example:
• No host pool tuning.
• No scaling logic to maintain.
• No infrastructure lifecycle to plan.

The tradeoff is that architects cannot intervene when edge cases arise; however, for many organizations, that is considered more of an advantage rather than a limitation.

Many struggling environments share a familiar pattern: the platform was chosen correctly for the use case, but the operating model was never adjusted. Common examples include:

  • AVD environments are deployed by project teams and then transitioned to IT operations teams that lack the skills and/or tools to manage image lifecycle and cost optimization.
  • W365 environments are expected to support highly customized workloads that require infrastructure-level tuning.
  • Organizations adopting AVD without clear ownership of autoscaling, cost governance, and/or image management.

In these scenarios, over time, issues typically manifest as:

  • Gradual performance degradation.
  • Rising, unexplained costs.
  • Security posture drift.
  • Increasing operational friction.

These are not platform failures; rather, they indicate day-2 ownership was never fully defined.

Note that W365 offers a number of “Hosted on behalf of” architecture models, effectively attaching Azure services to a hosted subscription. Depending upon the model appropriate for the organization, the Network team must be engaged to assist with implementing and managing the W365 footprint.

Operating Model Precedence

The most practical design question is often not “Which platform fits the use case?” but is instead “Which platform fits the way we actually operate today, and the way we expect to operate two years from now?” In this context, the operating model should guide the decision of which platform to use.

Organizations that succeed with AVD typically:

  • Have clear ownership for platform engineering tasks.
  • Treat cost optimization as ongoing work.
  • Accept complexity in exchange for control, or:
    • Employ AVD (or W365) add-on management tools such as Hydra or Nerdio.
    • Employ Infrastructure as Code/IaC (e.g., Terraform, Ansible, etc.) to automate deployment for AVD (or W365) workloads.
      • Note: IaC deployments can vastly simplify/ease efforts around workload management; however, the initial effort to implement an IaC solution is typically a very heavy lift and requires ongoing maintenance effort.

Organizations that succeed with W365 typically:

  • They are already mature in endpoint and identity management.
  • Value predictability over tunability.
  • Prefer service ownership over platform ownership.

If operational practices are not part of the decision-making process, even a technically sound design often becomes unsustainable. Architects and engineers should treat operating model alignment as a first-class design element rather than an afterthought, for example:

  • If the organization is prepared to operate a platform, AVD can be powerful and cost-effective.
  • If the organization prefers to consume a service, W365 will almost always age better.
  • If different teams or user populations operate differently, a mixed model often provides the most aligned and resilient design.

This distinction becomes increasingly important over time, because operating models tend to change more slowly than platforms. Platform choice should align with business requirements and should not conflict with operational practices and expectations.

Both AVD and W365 integrate into Microsoft’s broader management and security ecosystem, but their operational center of gravity differs:

  • AVD assumes stronger cloud infrastructure and operational maturity.
  • W365 shifts the focus toward identity, policy, and lifecycle management via Intune.

If an organization’s support structure cannot sustain the operational model, even a well-designed architecture will likely struggle as growth, shifts in security posture, and/or cost pressures occur. Many well-designed deployments fail because the operating model was not considered at the onset.

Decision Framework

By this point, the distinction between AVD and W365 should be clear: rather than competing models, they are delivery models optimized for different outcomes. The remaining challenge is turning that understanding into practical decisions that can be applied consistently.

The following table provides a simple framework to help architects and engineers make defensible choices quickly, without modeling every variable up front.

ScenarioRequirementsRecommended Platform
Always-On, Dedicated Desktops Per UserIf users require:
• Persistent, personal desktops.
• The ability to customize user environments.
• Minimal dependency upon shared infrastructure.
• Predictable availability regardless of time of day.
• AVD skillsets are not available or desired on the IT Operations team.
W365 is often the most appropriate starting point, as attempting to reproduce this experience with AVD often leads to over-provisioned host pools, reduced scaling efficiency, and higher operational overhead than expected.

Potential exceptions here would be influenced by:
• IaC requirements, which could be better suited to AVD.
• A preference to not be locked into a specific VM SKU after deployment (AVD would better accommodate the requirement for adjustments in VM SKUs).
• GPU requirements can be accommodated more economically with AVD versus W365.
• Machine provisioning will vary across platforms, with faster VM provisioning time offered by AVD with Hydra/Nerdio as compared to W365 and Intune.
High User Counts with Predictable ConcurrencyIf the environment involves:
• Large user populations with overlapping usage patterns.
• Similar application sets.
• Sensitivity to cost per user.
• Tolerance for shared infrastructure.
AVD is typically a better fit, particularly when multi-session desktops can be leveraged. In these scenarios, AVD’s additional design effort directly enables cost and density optimization.
Shift-Based or Non-Concurrent AccessIf users:
• Only require desktop access during scheduled shifts.
• Do not need concurrent access.
• Share workstations or roles.
W365 Frontline should be evaluated early. Frontline aligns entitlement with usage patterns rather than named users, reducing the need to force these workloads into either pooled VDI or always-on dedicated desktops.
Environments with Strong Cloud and Platform Engineering CapabilityIf the organization:
• Has cloud operational maturity.
• Is comfortable owning image lifecycle and scaling logic.
• Actively manages consumption and optimization.
AVD can provide long-term flexibility and cost control that service-based models intentionally abstract away.
Environments Optimized for Endpoint and Identity ManagementIf the organization:
• Is already standardized on Intune and Entra ID.
• Treats desktops similarly to physical endpoints.
• Prioritizes consistency and operational simplicity.
W365 will often integrate more naturally into existing processes and age more gracefully over time.

Consumption Cost v. Subscription Cost

While this blog has focused on use cases and design intent rather than cost optimization, it is still important to acknowledge that AVD and W365 follow fundamentally different cost models, and this distinction alone often influences architectural decisions.

AVD is primarily a consumption-based model. Beyond licensing eligibility, organizations incur costs based on Azure resources consumed, e.g., compute, storage, networking, and optional services. This model rewards environments that are well-architected, right-sized, and actively optimized, particularly where concurrency patterns allow for autoscaling and efficient host utilization.

Alternatively, W365 is a subscription-based model. Each Cloud PC (or Frontline entitlement) entails a predictable per-user monthly cost that bundles infrastructure, platform operations, and availability into a single charge. This predictability is often valued as much for budgeting and financial planning as for operational simplicity.

This becomes significant at the break-even point. In practice, organizations often find that:

  • Always-on, persistent, per-user desktops tend to favor W365 from a predictability and planning perspective.
  • Variable, bursty, or non-concurrent workloads can favor AVD, where consumption can be actively controlled.
  • Shift-based and intermittent access scenarios increasingly align well with W365 Frontline, which changes the economics by aligning licensing to usage patterns (rather than to named users).

Importantly, cost outcomes are rarely driven by platform choice alone. They are the result of how closely the delivery model aligns with actual usage, how effectively environments are operated over time, and whether the organization is prepared to manage the cost levers exposed by the selected model.

A Note on Cost Models and Break-Even Considerations

Although this blog intentionally focuses on use cases rather than cost comparison, it is important to recognize that cost behavior follows design intent.

  • AVD’s consumption-based model rewards environments with well-understood concurrency and disciplined operational practices.
  • W365’s subscription-based model favors predictability, particularly for always-on or dedicated user scenarios.
  • W365 Frontline introduces a different break-even point by aligning costs to non-concurrent usage.

In practice, cost outcomes are typically the result of how closely the delivery model aligns with real-world usage and how effectively the environment is operated over time.

Closing Thoughts: Designing for Sustainability, Not Just Deployment

AVD and W365 both solve real problems and do so best when applied to the right use cases, with the right expectations, and under the right operating model.

This use-case-driven architecture/design model has been reinforced through real-world design workshops and community working sessions, including discussions with Microsoft MVP Thomas Poppelgaard, whose expert guidance on positioning AVD and W365 as intentionally distinct tools continues to resonate in practice.

Ferroque Systems has the expertise and experience to guide organizations in determining the best-fit virtual delivery solution, including with AVD and W365, and to design, implement, and roll out such solutions, as well as to provide ongoing operational support and management. We would be happy to discuss how we can help.

 

  • Jeremy Ingram

    Jeremy has been deploying Citrix and NetScaler products since 2008. A seasoned architect spanning technologies and industries, Jeremy has a passion for deploying Citrix products, which he firmly believes are the coolest bits running in enterprise environments today.

Subscribe
Notify of
guest

0 Comments
Inline Feedbacks
View all comments

Redefine Your Approach to Technology and Innovation

Schedule a call to discover how customized solutions crafted for your success can drive exceptional outcomes, with Ferroque as your strategic ally.