Intelligent Device Refresh with Nexthink: A Step-by-Step Implementation Guide

For myself and most of the organizations I've worked with, intelligent device refresh was a concept that looked better on a whiteboard than it did in practice. The effort to build a data-driven refresh program rarely justified the savings. 

That math has changed.

AI adoption is pushing demand for higher-spec endpoints. Semiconductor constraints have tightened supply and pushed prices up. Lower-end devices are harder to source, which means organizations are now being forced into more expensive hardware at every turn. 

Intelligent device refresh went from being more of a marketing point, to something being looked at on a weekly basis as organizations try to figure out how to get smarter about refresh before the next budget cycle.

What intelligent device refresh actually means

Instead of replacing devices on a calendar, you replace them based on how they're performing for the people using them. Digital Employee Experience (DEX) data from Nexthink gives you continuous telemetry across hardware, applications, OS and network, translated into an experience score that tells you whether a device is delivering or degrading.

The result is a three-way decision instead of a binary one:

• Keep: Device is performing well. Extend the lifecycle and defer the spend.
• Upgrade: Underperformance is isolated to memory or disk. Where possible, a targeted component swap can extend the device's useful life.
• Replace: Overall device is underperforming to user needs. Replacement is justified and defensible.

How to be successful with intelligent device refresh

The following scenarios reflect patterns we see repeatedly across organizations navigating managing device refreshes in today's reality.

Budget crisis forces a smarter approach. Per-unit costs are climbing and price quotes are good for weeks at best, meaning the device count approved last budget cycle no longer covers what was planned. Organizations using Nexthink cut the refresh list to what actually needs to go, what can run longer, and what needs a targeted upgrade rather than a full replacement. The refresh gets completed closer to budget without a measurable drop in employee experience.

Ever-increasing hardware requirements. Most organizations are finding that their base endpoint spec is what used to be their developer class just a few years ago. Rather than simply provisioning more hardware, Nexthink DEX data points directly to the performance logs, software settings and drivers worth optimizing, and agent consolidation opportunities that can meaningfully reduce what the next refresh actually needs to cost.

Flying blind on forecast. Once you move off a fixed refresh cycle, how do you plan how many devices of each type you'll need next year? Actual performance data shows what hardware is really being consumed, so device specs are sized to actual use rather than assumption. Trending that data forward turns budget forecasting from a guess into a defensible projection.

The Nexthink intelligent device refresh playbook

Here's how to get started with Nexthink, from licensing through your first analysis.

Step 1: Confirm your licensing

To run intelligent device refresh in Nexthink, you need two modules licensed:

• Workplace Experience — core endpoint monitoring across hardware, applications and network
• Digital Experience — the DEX scoring engine that ties device performance to user experience

The DEX score is the mechanism that tells you, with data, whether a device is worth keeping, worth upgrading or ready to replace. Without both modules active, the Hardware Refresh Insights dashboard won't have the scoring data it needs.

Step 2: Install the Hardware Refresh Insights library pack

From the main menu in Nexthink Infinity, navigate to Library and search for "Hardware refresh insights." Install the pack. It deploys four items:

• Hardware Refresh Insights dashboard — the primary analysis surface, organized around DEX scores
• Get warranty information — a remote action that pulls warranty status from Dell, HP and Lenovo via their manufacturer APIs
• Get battery status — a remote action that returns battery health for laptops, feeding into the upgrade analysis tab
• Hardware refresh review — a custom field that flags which devices are in scope for analysis

The dashboard only analyzes devices where the Hardware Refresh Review field is set to "Under review," so nothing runs against your full fleet until you tell it to.

Step 3: Configure the remote actions

Navigate to Remote Actions, then Manage Remote Actions, and locate the Get warranty information and the Get battery status actions from the library pack.

If your devices are from Dell, Lenovo or HP, you can configure the Get warranty information by entering the API credentials obtained from each manufacturer's developer portal. Once configured, run this action manually against your device fleet to pull current warranty status. Out-of-warranty devices are a natural starting point for your candidate list.

If your environment doesn't include these vendors, or you don't have API credentials readily available, you can still proceed to Step 4 using one of the other candidate list options.

For Get battery status, the only parameter is temperature unit: Celsius or Fahrenheit. Set it and it's ready. This action feeds battery health data into the Devices to Upgrade tab.

Both actions are set to manual trigger by default. You control when they run and against which devices.

Step 4: Build your candidate list

The dashboard only scores devices tagged "Under review," so this step determines the scope of your analysis. A few practical approaches:

Option 1: Use the warranty remote action - Run it across your fleet, identify out-of-warranty devices and tag them. Warranty expiration correlates well with age and is a defensible starting point.

Option 2: Manually upload device list — From your IT asset management (ITAM) you can export a list of devices at or near end of life. Create a CSV with the devices listed and upload into Nexthink via the Custom Fields page to tag devices in bulk

Option 3: Build a custom automated list — Using Nexthink services or partner, you can build connectors to ITAM systems for automated imports and tagging.

Regardless of method, the goal is the same: get the right devices in scope before you open the dashboard so you're analyzing a meaningful population, not your entire fleet.

Step 5: Run the analysis

With your candidate devices tagged, open the Hardware Refresh Insights dashboard. It's organized into four tabs.

Device Selection — This is your starting point. It shows how many devices are currently tagged for review out of your total active population, broken down by model with average DEX scores per model. It also surfaces any devices outside your tagged population that are scoring below 30 on the DEX, giving you a safety net for devices that should have been in scope but weren't.

Devices to Replace — Shows devices underperforming across multiple DEX components, not just memory or disk. These don't qualify for a targeted upgrade because the issues are too widespread. Drill into component scores here to build a prioritized replacement list grounded in experience data rather than age.

Devices to Upgrade — Identifies devices where overall DEX is dragging, but the problem is isolated to memory or disk. CPU, network and application scores are satisfactory; memory or disk is not. These are your targeted upgrade candidates. 

Devices to Keep — Shows in-scope devices scoring above 70. Even if these devices are out of warranty or over five years old, the data supports extending them. Drill into the individual DEX component scores to confirm no single area is being masked by a strong overall score. If everything checks out, you have a defensible case for deferring replacement.

Step 6: Automate the refresh lifecycle

The Hardware Refresh Insights library pack gets you to first analysis quickly, but it has a ceiling. For organizations that want to go further, Nexthink supports fully custom dashboards and workflows that automate ongoing device health analysis, forecast how many devices will need replacement in upcoming quarters, and integrate directly with ITSM platforms to trigger refresh workflows automatically.

This isn't an out-of-the-box configuration. It requires custom development, either internally or through Nexthink Professional Services. But for organizations managing large fleets where refresh planning is a continuous operational function rather than an annual exercise, it turns the one-time analysis in Steps 1 through 5 into a repeatable, automated operating model.

Beyond refresh: Getting more out of what you already have

Identifying which devices to replace is only part of the equation. Before cutting a PO, ask whether the device's problems are actually hardware problems. Nexthink telemetry tells you why a device is underperforming, and that distinction determines whether the right answer is a configuration change, a component swap, or a full replacement. A few areas worth evaluating before you pull the trigger on a replacement order:

• Right-size new devices using DEX data. Rather than overprovisioning "just in case," use actual CPU utilization, memory pressure, and storage consumption data to build defensible user personas. Size new hardware to what users actually need, not what IT assumes they need.
• Optimize existing devices before you replace them. OS-level tuning using tools like WDOT can reduce unnecessary background service consumption on physical endpoints, not just VDI. Small configuration changes identified through DEX analysis can meaningfully improve device performance without requiring hardware changes.
• Reduce agent sprawl. EDR, VPN clients, UEM agents, patching tools and asset inventory running simultaneously add up fast. Consolidating overlapping tools and tuning agent behavior, shifting scans off-hours, reducing telemetry frequency, and moving to a split-tunnel VPN, can reclaim CPU and memory headroom without compromising security or manageability.

Calendar refresh is no longer a viable strategy

For most of my career, intelligent device refresh wasn't worth the effort. It is now. 

The prices are too high, the availability too unpredictable, and the tooling too good to keep making refresh decisions with a calendar.

Start with your out-of-warranty population and run the analysis. The data will take it from there.

If you want a guided conversation on where to start, WWT offers a Digital Employee Experience briefing tailored to your environment and fleet size.