Mass migration to the next device landscape

Manufacturer makes short work of a large-scale Windows upgrade migration

At a glance

A company’s 300,000 end-user devices embarked on a data-driven migration to Windows version 20H2. Many were not ready. Many were considered incompatible. Even so, the upgrade was completed in six months with surprising results.


  • Record-low failure rate of 0.55%.
  • 21% fewer devices determined incompatible.
  • Tens of thousands of user productivity hours saved.

Dreaded chores

A global manufacturing organization planned for a six-month Windows upgrade migration to version 20H2 across 300,000 devices. The challenge was that many of those devices were not considered ready to migrate, requiring the IT service teams to conduct endless manual compatibility and compliance checks to try and avoid failures during the process.

A Windows migration is like putting off a chore you really hate doing. You can delay it for a long time, but you have to do it at some point. And you’ll be so happy once it’s done.Unlike delaying the chores, however, the risk of failure throughout the entire migration process is extremely high, typically around 15% of Windows upgrades like this one fail.

In fact, this is exactly what happened during the previous Windows update to version 1909. It occupied many resources with weeks of challenges due to upgrade failures. Thousands of hours of productivity were lost, and employees were frustrated. To avoid recreating that situation, the IT services teams wanted to do things differently this time. They looked to use the proactive data-driven services from Atos to make the migration easier.

Next-level prepping

Before the migration, the End-User Computing Analytics (EUCA) team at Atos used Nexthink’s digital employee experience platform to conduct readiness assessments of the manufacturer’s hardware and software environments. They remotely identified every element that could potentially prevent successful upgrades. Then they used the platform’s contextual visibility, automation and targeted engagement features to resolve the issues.

(More information about the Nexthink platform.)

Hardware readiness actions

Ensure disk space

Problem: Many devices could not be upgraded due to a simple lack of disk space.

Solution: Analytics & Self Help.
Automated identification of every device with low disk space followed by deployment of targeted employee engagement campaigns asking employees if they wanted an instant, automatic disk clean-up with one click.

Outcome: 4,140 devices were cleaned, saving hundreds of IT support hours.

Verify device model compatibility

Problem: Some specific device models simply couldn’t be upgraded, but the landscape could not be filtered by device model to allow effective engagement with the right users.

Solution: Analytics & User Engage.
Search and filter results of all incompatible devices, then engage with employees to automatically trigger tickets and schedule a device replacement or hardware upgrade — all via Nexthink.

Outcome: Over 5,000 devices were automatically scheduled for upgrade with user consent, saving over 1,000 IT support hours from potentially failed upgrades.

Identify BIOS non-compliance

Problem: The Windows upgrade required a specific BIOS setting, but the EUCA team did not know which devices had it enabled.

Solution: Analytics. Use the platform to identify devices that did not have the required BIOS and schedule them for the necessary upgrade.

Outcome: Over 4,000 devices were scheduled for upgrade, saving over 500 IT support hours from potentially failed upgrades.

Update OS build

Problem: All devices with an OS build older than 17763.1158.0.9 were incompatible, but the team had no way to specifically find, filter and target these devices all at once.

Solution: Analytics & Remote Action. Nexthink’s instant visibility into OS versioning non-compliance allowed the IT service teams to target and upgrade all devices remotely.

Outcome: Device incompatibility due to OS non-compliance was reduced by 72% in a single action.

Detect pending reboot

Problem: Devices could not be properly upgraded without a full reboot and were at risk of migration failure.

Solution: Analytics. Identify every device with a pending reboot, via an investigation in Nexthink, and deploy targeted engagement with those employees to reboot their devices.

Outcome: 120,000 devices were identified for reboot via a campaign pop-up asking the user to reboot.

Driver & BIOS registry

Problem: Devices without specific driver or BIOS registry setting were not upgradable.

Solution: Analytics & Remote action. Investigate the registry setting of driver and BIOS to identify devices that needed a registry update before they could be upgraded.

Outcome: 50 work-hours were saved from failed upgrades due to incompatible driver registry settings, and 6,000 devices were identified with the wrong BIOS registry settings.

Software readiness actions

VMware and VirtualBox

Problem: The IT service teams knew that devices with VMWare Workstations or VirtualBox could not be successfully upgraded, but many users would have installed those applications offline, so IT had no real visibility of their inventory.

Solution: Analytics. Instantly investigate via Nexthink investigation every device with VMWare and/or VirtualBox installed across the entire landscape.

Outcome: Over 100 IT support hours were saved from failed upgrades due to the presence of these applications.

Update SCCM

Problem: The IT service teams noticed that devices with an SCCM version older than 5.00.9040.1044 would result in a failed upgrade.

Solution: Analytics & Remote Action.
Investigate SCCM versioning across the device landscape via investigation and send targeted, automated remediation to update devices with outdated versions.

Outcome: Over 5,000 devices were automatically scheduled for upgrade with user consent, saving over 1,000 IT support hours from potentially failed upgrades.

Identify BIOS non-compliance

Problem: The Windows upgrade required a specific BIOS setting, but the EUCA team did not know which devices had it enabled.

Solution: Analytics. Use the platform to identify devices that did not have the required BIOS and schedule them for the necessary upgrade.

Outcome: Over 1,000 IT support hours were saved from failed upgrades with a 77% decrease in non-compliant devices.

Enable CrowdStrike agent

Problem:A CrowdStrike agent must be present on a device for a successful upgrade, but those are often uninstalled, broken or disabled.

Solution: Analytics. Identify devices with non-compliant CrowdStrike agents and remotely update them.

Outcome: Over 1,000 IT support hours were saved from upgrade failure with a 93% reduction in non-compliant devices

Ensure Windows Defender compliance

Problem: Windows Defender RTP status and updated status must be enabled for a successful upgrade, but manually checking for these statuses was a tedious manual process.

Solution: Analytics & Remote Action. Immediately identify devices with a non-compliant Windows Defender and deploy an automated configuration upgrade across all of them at once.

Outcome: Saved over 5,000 IT support hours from upgrade failures with a 72% reduction in non-compliant devices.

Remove blacklisted applications

Problem: The company put together a blacklist of applications that, although might not prevent a successful upgrade, could represent future security risks and must be uninstalled. But its IT service teams had no visibility into shadow IT activity or how to prevent it at scale.

Solution: Analytics. Investigate and list every device with any blacklisted applications. The IT service teams could then deploy engagement with users and trigger automated uninstallations across all concerned devices.

Outcome: Nearly 10,000 IT support hours were saved with 80% reduction in devices representing security risks.

How low can your failure rate go?

Working together, the manufacturer and Atos were able to resolve all of the issues in only a matter of days.

Those devices that were previously considered incompatible? The readiness assessment and resolutions reduced the total number by 21% and saved tens of thousands of user productivity hours.

What’s more, the Atos EUCA service, powered by Nexthink, resulted in only a 0.55% upgrade failure rate – the lowest the client had ever seen in projects of this scale and complexity.

Apr 21 – May 21


June 21 – Aug 21


Aug 21 – May 22


May 22


Analyze, assess, scope and plan.

Creation and testing of solutions for HW and SW readiness.

Schedule and run remediation actions as well as provide reporting to kick start migration.

Close of project followed by Closure Report containing full details for future reference.

Why Atos

End-User Computing Analytics (EUCA) services from Atos deliver a can-do culture at the heart of proactive support, with greater engagement, feedback capture and action based on real-time user experiences working with existing service models.

Related resources


Frictionless working: How to thrive in the new, blended world of work

Only by simplifying communication, collaboration and knowledge sharing will individuals and teams be empowered to grow.


Voice of the Employee – Make sure the feeling is right

Employee services should not only provide a quick and solid solution, but a positive employee experience at the same time.


Leave productivity to the robots, it’s time to care for employees

The onus of productivity must not be on your employees: put simply, they’re not machines, they’re human beings.

More client stories

Interested in our solutions?