How to optimize your end-user computing environment to escape the AI price increases

In my last article, I explored why adopting an intelligent device refresh strategy is becoming a strategic imperative, particularly in light of rising prices and ongoing availability challenges for end‑user devices. Rather than refreshing devices solely based on age, an intelligent refresh strategy replaces devices based on performance, user experience, role, and workload. This approach can significantly reduce capital expenditure, improve sustainability, and provide greater resilience during endpoint shortages.

These benefits are amplified when the strategy is supported by experience data from a Digital Employee Experience (DEX) solution. DEX insights help identify the right users and devices, where devices can be extended safely without compromising productivity.

However, many organizations are not in a position to fully, or even partially, adopt an intelligent refresh strategy. Financial and accounting constraints, the operational and support changes required, or the simple reality that existing devices are already underpowered for current and future workloads, can all limit adoption.

The good news is that meaningful gains in performance, stability, and cost efficiency are still achievable, even within your existing endpoint estate and refresh model. By focusing on software efficiency, workload placement, and data‑driven decision making, IT teams can improve employee productivity while keeping costs and complexity under control. And even better news is that all of these strategies also apply to organizations changing to an intelligent refresh model!

Here are six strategies to help you get the most out of your end‑user computing environment:

1. Right-size devices using DEX to avoid overprovisioning

Last time, we discussed how DEX data can enable an intelligent refresh strategy. To build on that, let's start by examining how DEX can also add value within a more traditional device refresh approach.

Historically, a common strategy when ordering new devices was to overprovision hardware "just in case." In today's reality, where device prices are rising and availability is constrained by demand for AI‑capable components, this approach is no longer sustainable. DEX data gives IT teams the insight they need to make more precise, defensible decisions.

By analyzing real‑world usage metrics captured by a DEX platform, organizations can develop a far more accurate understanding of their users' actual requirements. Beyond high‑level CPU utilization, memory pressure, and storage consumption, IT teams can also examine the impact of individual applications and management agents. With this level of insight, new endpoint hardware can be sized appropriately, reducing unnecessary overprovisioning through data‑driven user personas.

DEX insights also improve forecasting accuracy. Trends in increasing resource requirements can be identified early and projected across the expected service life of new devices. Once those devices are deployed, ongoing monitoring helps ensure that changes such as operating system updates, new applications, or firmware revisions do not materially alter the assumptions made during procurement.

Continuous review of DEX findings allows IT teams to quickly identify poorly performing applications or suboptimal operating system configurations. Even small changes uncovered through this analysis can have a meaningful impact on device performance and overall employee experience.

2. Reduce agent sprawl to reclaim performance

Most enterprise endpoints today run a surprising number of background agents: endpoint detection and response (EDR), VPN clients, device management, patching tools, asset inventory, digital experience monitoring, collaboration software helpers, and more. Individually, each agent may appear lightweight, but collectively they can consume significant CPU, memory, disk I/O, and battery life, particularly on older devices.

There are two primary strategies for reducing the cumulative impact these management agents have on overall device performance:

There are two main strategies for reducing the impact of these management agents that are impacting the overall device performance:

Consolidate - One of the most common optimization opportunities is consolidating overlapping tools. Many organizations still deploy separate agents for endpoint protection, vulnerability scanning, and device inventory, even though modern unified endpoint management (UEM) or security platforms can deliver multiple capabilities through a single agent. Reducing three or four persistent services down to one or two can immediately lower idle CPU utilization, decrease memory pressure, and improve boot and login times without reducing functional coverage.
Optimize - Another effective approach is tuning agent behavior. EDR scans scheduled during business hours, overly aggressive telemetry collection, or continuously active "always‑on" VPN tunnels can significantly degrade the user experience. Simple adjustments such as reducing scan frequency, shifting resource‑intensive tasks to off‑hours, or moving from full‑tunnel to split‑tunnel VPN configurations can materially reduce device strain without compromising security or compliance.

By consolidating redundant tools and optimizing agent behavior, organizations can reclaim valuable CPU and memory headroom. The result is improved performance on existing or lower spec hardware, better employee experience, and all without sacrificing security or manageability.

3. Enhance endpoint performance through resource optimization

The next step is to get more out of the devices you already own before facing costly hardware upgrades. This comes from tuning every layer of the endpoint stack, from the operating system and applications to how hardware resources are consumed under real‑world usage.

Historically, operating system tuning has been most common in Virtual Desktop Infrastructure (VDI) environments, where even small efficiency gains translated directly into reduced hypervisor costs. However, many of these same optimizations are equally applicable to physical endpoints, particularly as devices age or are pushed closer to their performance limits.

Modern OS optimization tools analyze system configuration, background services, and default settings, then recommend changes that reduce unnecessary CPU and memory consumption. Most provide a review or "analyze" mode that allows administrators to validate recommended settings before applying them. As with any performance tuning, changes should be deployed selectively and continuously monitored using a Digital Employee Experience (DEX) platform to confirm that optimizations improve responsiveness rather than introduce instability.

Examples of commonly used OS optimization tools include:

Windows Desktop Optimization Tool (WDOT)
Omnissa OS Optimization Tool (active Omnissa agreement required)
Citrix Optimizer Tool (active Citrix agreement required)

Beyond OS‑level tuning, additional gains can be achieved by managing how applications and background processes consume resources during periods of peak demand. Process optimization tools monitor CPU and memory usage in real time and detect when background applications or services are consuming disproportionate resources. When this occurs, they dynamically adjust CPU priority or reclaim memory from non‑active processes, ensuring that foreground applications remain responsive.

This approach helps smooth performance spikes, reduce contention, and improve the perceived responsiveness of the system without requiring changes to the applications themselves.

Examples of process‑level optimization tools include:

Liquidware Stratusphere UX Process Optimization
Citrix Workspace Environment Manager (WEM) (active Citrix agreement required)
ControlUp (using real‑time triggers + automated actions)
Process Lasso

When guided by DEX insights, these optimizations help IT teams extract maximum value from existing hardware before investing in higher-spec hardware.

4. Modernize application delivery to reduce device demands

Thick, device‑installed applications are often resource‑intensive and tightly coupled to the operating system. Where feasible, migrating these applications to web‑based or Software-as-a-Service (SaaS) alternatives can materially reduce endpoint resource consumption and operational overhead.

A classic example is moving from a locally installed CRM client to a browser‑based SaaS platform. Instead of maintaining background services, local client storage, and frequent client updates, the endpoint simply runs a modern web browser. This approach not only reduces the load on the device but also simplifies patching, upgrades, and compatibility management.

Native applications can further complicate operating system upgrades and patching, as they must be rigorously tested for compatibility and continuously monitored for performance impacts. With web‑based applications, IT teams can standardize on a supported browser and manage access centrally. For many users, perceived performance actually improves, as SaaS platforms are optimized for distributed, cloud‑scale execution rather than local processing.

Shifting from device‑native software to web and SaaS applications offloads CPU, memory, and storage requirements from the endpoint to the cloud. The result is lower resource demand on existing hardware, improved consistency across users, and a simpler, more manageable endpoint environment.

5. Move the workload, not the hardware

When device‑level optimization reaches its practical limits, the next lever to consider is moving the workload itself. Cloud‑hosted virtual desktops and applications enable organizations to deliver high‑performance user experiences on virtually any endpoint. By migrating users to a Cloud PC or virtual application, the local device effectively becomes a secure access terminal, with significantly reduced resource requirements.

Additionally, as the cost of physical endpoint devices continues to rise, migrating desktops and applications to the cloud is becoming both financially and operationally attractive. Hyperscale cloud providers leverage locked‑in pricing and economies of scale to mitigate the impact of rising hardware costs relative to on‑premises environments. While consistent, on‑premises VDI deployments were historically more cost‑efficient, the cost differential between on‑premises and cloud‑hosted desktops is now smaller than it has ever been.

Cloud‑hosted virtual desktops and applications now support a broader range of use cases than ever before, including contractors, back‑office staff, front‑line and shift workers, disaster recovery scenarios, and even users with GPU‑accelerated workloads. They are particularly well suited for external 3rd-party users, task workers, and in regulated environments where data must remain centralized and tightly controlled.

Organizations can selectively migrate targeted use cases to the cloud using solutions such as:

By leveraging Cloud PCs and virtual applications, organizations can decouple the user experience from local hardware constraints, delivering the required performance and security without costly physical device refreshes.

6. Rethink the endpoint operating system

A final option is moving away from Windows on endpoints in favor of Linux. While this may initially sound more aspirational than practical, it is increasingly viable for a meaningful subset of users and use cases.

Consider this scenario: if all applications used by a group of users are either web‑based or delivered through virtual applications or desktops, the underlying endpoint operating system becomes far less critical. In these cases, endpoints can be readily converted to Linux. This model has existed for years in VDI environments, but today it applies to a broader set of users than ever before.

For example, endpoints converted to Linux‑based IGEL OS can support:

Native web browsers - Chromium, Edge, and Firefox
Enterprise browsers - Island and Prisma
Progressive Web Apps (PWAs) - Microsoft Excel, Word, Outlook and more
Virtual application and desktop access - Omnissa, Citrix, and Microsoft platforms

Not only can Linux improve endpoint security and reliability, but it is also more resource efficient than running Windows on the same hardware. CPUs that may not meet Windows 11's minimum requirements are often more than sufficient for Linux workloads. Memory efficiency is also higher, as Linux does not reserve large portions of system RAM for integrated graphics' virtual video memory as Windows does.

As part of an updated persona‑based analysis, organizations should evaluate whether certain user groups could transition to a different endpoint operating system based on their actual requirements. For the right users, a Linux‑based endpoint can extend hardware life, reduce resource pressure, and simplify endpoint management.

Become more intentional with endpoint strategy

Organizations can materially improve performance, stability, and cost efficiency without defaulting to expensive hardware upgrades by right‑sizing devices with DEX data, consolidating and tuning management agents, optimizing CPU and memory utilization, shifting applications to web and SaaS platforms, moving targeted users to cloud‑hosted desktops, and rethinking the endpoint operating system itself.

These strategies share a common principle: Optimize before you replace. Rather than relying on brute‑force refresh cycles, successful teams focus on efficiency, insight, and workload placement. The result is extended device lifecycles, better employee experiences, and greater flexibility in the face of constrained budgets and volatile hardware markets.

Now is the time to take a more intentional approach to end‑user computing. Start by understanding how your devices are actually used, identify where optimization can deliver immediate value, and apply change selectively and where it has the greatest impact. Whether you are working toward an intelligent refresh strategy or simply buying time before the next refresh cycle, these steps help IT leaders regain control of cost, performance, and user experience in an increasingly AI‑driven world.