Digital employee experience for Microsoft 365: what to measure

Digital employee experience (DEX) is how well the technology people use every day works for them, measured from their side. For Microsoft 365, it means measuring how Outlook, Teams, SharePoint and OneDrive respond on each device and network, how calls and meetings perform, and how often things fail, continuously rather than after a complaint.

Someone in IT usually has to turn "improve the employee experience" into figures the helpdesk can act on. This guide lists the indicators that matter for Microsoft 365, what Microsoft already gives you, and how we approach user experience monitoring by app and platform.

What is digital employee experience for Microsoft 365?

It is the part of the employee experience that IT can measure: whether the tools work, quickly, wherever people sit. Surveys tell you how people feel about their tools. Measurement tells you what they actually got: the sign-in that took a minute, the Outlook search that hung, the call that froze in front of a customer.

For most organisations, Microsoft 365 is where the working day happens. Outlook, Teams, SharePoint and OneDrive stay open from morning to evening, on laptops, phones, virtual desktops and the web. A DEX programme that leaves them out measures the wrong thing. One that only watches the devices misses the network and the service in between.

Why does a good overall score hide the people who struggle?

Because problems rarely hit everyone at once. An organisation-wide score can look healthy while one office, one ISP, one laptop model or one virtual desktop pool has a poor day every day. Those people stop reporting it and find workarounds.

Two habits help:

• Break every indicator down by site, ISP, device model, platform and app, so a small group with a large problem stands out.
• Compare each person with their own normal, so a sudden drop stands out from a long-standing problem.

For calls, the Teams admin center already has user, meeting and participant views, with real-time telemetry for running meetings (Monitor and troubleshoot Teams meetings and calls). Our guide to Teams and Zoom call quality monitoring explains how to trace one poor call to its cause.

What should you measure for digital employee experience?

Measure each layer between the person and the service, because each one can fail on its own: the device and its resources, the local network and Wi-Fi, the headset and how it is connected, the ISP, the VPN, sign-in, then the Microsoft 365 apps themselves. The table lists the indicators we find most useful, and where the data comes from.

Measure continuously, not only after a complaint, so there is always a "before" to compare with, and keep the history to check that fixes hold.

What does Microsoft provide natively?

More than many admins use. The tools sit in Intune, the Microsoft 365 admin center and the Teams admin center, and each one answers part of the question.

• Endpoint analytics (Intune). Scores startup performance, application reliability and remote work readiness, from 0 to 100. Application reliability ranks desktop apps by crashes over 14 days. The Work from anywhere score measures how ready devices are for remote work, not how a session felt. Endpoint analytics covers managed Windows devices and needs an Intune licence (Endpoint analytics overview).
• Intune Advanced Analytics. Adds battery health, CPU and memory performance, anomaly detection and a device timeline. It is now included in Microsoft 365 E3 and E5, with a rollout Microsoft set to complete by 1 August 2026, and sold as an add-on otherwise (Microsoft Intune Advanced Analytics, Microsoft 365 packaging and pricing updates).
• Network connectivity (Microsoft 365 admin center). Scores how well each office location connects to Microsoft 365, from 0 to 100, once offices are declared or discovered. Remote workers are reported by city, not by person (Network connectivity in the Microsoft 365 admin center).
• Usage reports. Show who uses what, with a delay of one to three days, and hide names by default (Microsoft 365 usage reports).
• Teams admin center and Call Quality Dashboard. Live participant data for running meetings, and call and meeting quality trends across the organisation, by network, building and site (What is Call Quality Dashboard). With Teams Premium, the Teams admin center can alert on audio, video and screen sharing quality in running meetings, for the users you choose (Alerts in the Teams admin center).
• Experience insights. A dashboard that combines usage with user feedback and NPS, optimised for organisations with 2,000 licences or more and still in preview on 6 October 2026 (Experience insights dashboard).

One change matters if you follow older guides. Between 22 January and February 2026, Microsoft retired the Technology experiences score of Adoption Score, with its three categories: network connectivity, Microsoft 365 Apps health and Endpoint analytics. Adoption Score now measures how people use Microsoft 365 and Copilot, at organisation level only, not device or network health (Microsoft Adoption Score).

These tools are the right place to start. What remains is joining them up: a crash-prone Windows laptop shows up in Intune, a poor meeting at a branch office in the Teams admin center and a weak office connection in the Microsoft 365 admin center, each with its own scope and its own delay.

How do synthetic tests and real user data fit together?

They answer different questions, so neither replaces the other. Synthetic tests repeat everyday actions, such as signing in, sending a mail or opening a SharePoint page, from the sites you choose, around the clock. They tell you whether the service works from here, right now, even when nobody is at their desk. Real user data comes from a light agent on each device and tells you what this person got, on this device and this network, while they worked.

Microsoft uses the first method itself: its Exchange Online monitoring already uses synthetic connections from your devices (Exchange Online monitoring for Microsoft 365). Synthetic tests do not see a home router; real user data only covers the devices where the agent runs. Most organisations use both. Our article on SharePoint Online monitoring with synthetic tests and real user data compares the two methods side by side.

What we see in the field

Over 15 years of monitoring projects with large enterprises, first on-premises and then in the cloud, a few patterns keep coming back.

Real time forgives nothing. In Teams, people need to see each other, hear each other, share a screen and a file, quickly and from several places at once, and remote work has made that the norm. A mail that arrives a few seconds late goes unnoticed. A frozen call does not.

Teams has become a full platform. It is no longer only for calls, and when a meeting does not start, the cost is immediate. We have seen meetings where 10 to 15 people waited several minutes for the meeting to start. Fifteen people waiting ten minutes is 150 minutes, two and a half hours of working time lost for a single meeting. Apply your own hourly cost, then count how often it happens.

Many failure points sit on the user's side. The device and its resources, the local network, the headset (certified or not, and how it is connected) and the ISP each break calls and apps in their own way. That is why the table above starts at the device, not at the service.

Troubleshooting a senior manager's meeting while it runs is a subject of its own, covered in our Teams call quality guide.

How do you turn digital employee experience data into fewer tickets?

By sending each issue to someone who can fix it, and by working on the causes that cost people the most time. A practical order:

1. Start with a scope. Pick the apps, sites and user groups that matter most, often Teams and Outlook for a pilot group, and agree who receives which alert.
2. Measure a baseline first. Record each person's normal experience before setting thresholds, so alerts compare like with like.
3. Give each layer an owner. Decide before the first alert who acts on a device, headset, Wi-Fi, VPN, ISP or Microsoft issue, so no alert lands in a shared queue with no name on it.
4. Start each callback from the user's timeline. Device, network and app data side by side show which layer changed, and when, before anyone asks the user to restart.
5. Rank recurring causes by time lost. The sites, ISPs, device models and app versions that cost people the most minutes go to the top of the list.
6. Use device data in refresh decisions. Replace the laptops that measurably hold people back, not simply the oldest ones.
7. Check that each fix held. Watch the trend for that site or group over the following weeks before you close the problem.

Steps 5 to 7 are where tickets go down: a cause fixed at the source stops generating calls.

How Neptune helps

Neptune is an integrator. No single tool covers every layer equally well, so we select the experience monitoring tools that fit your devices and sites, implement them in your tenant with your team, train your helpdesk, workplace and network teams to read them, and support you during the project, then through a customer success manager. See why we work with several tools.

The tools we put in place cover the indicators above:

• App by app, against each person's normal. Each workload gets an experience score measured against that person's own normal, so a sudden drop stands out from a long-standing problem.
• Every Teams call as it runs. Each call is followed while it runs, one-to-one, group and PSTN calls included, on Windows and macOS, with fresh figures every 30 seconds. Device load, headset, Wi-Fi, VPN and ISP are recorded with the call, so the same detail is there live and days later.
• The device. CPU and memory under load are measured on Windows and macOS devices, and the same agent runs on virtual desktops such as Citrix and Azure Virtual Desktop.
• The network. Every hop from the device to Microsoft's network is measured, with 90 days of history, including whether the VPN split tunnel works as intended. ISPs and managed networks are ranked by the experience they deliver.
• The right owner. Each layer, from the device to Microsoft, is measured on its own, so an issue goes straight to the team or provider that owns it.
• Synthetic tests. Virtual users test Microsoft 365 from each site, around the clock, and raise an alert when a service slows down.
• Deployment. A light agent runs in the user's context, without admin rights, and is deployed with Intune or your usual tooling.

We agree with you which apps, sites and user groups to watch first, measure each user's baseline, then set thresholds and alert recipients with your helpdesk, network and NOC teams. See how we approach Microsoft 365 monitoring.

Sources

• Endpoint analytics overview (Microsoft Learn)
• Microsoft Intune Advanced Analytics (Microsoft Learn)
• Microsoft 365 packaging and pricing updates (Microsoft)
• Network connectivity in the Microsoft 365 admin center (Microsoft Learn)
• Microsoft 365 usage reports (Microsoft Learn)
• Experience insights dashboard (Microsoft Learn)
• Microsoft Adoption Score (Microsoft Learn)
• Monitor and troubleshoot Teams meetings and calls (Microsoft Learn)
• What is Call Quality Dashboard (Microsoft Learn)
• Alerts in the Teams admin center (Microsoft Learn)
• Exchange Online monitoring for Microsoft 365 (Microsoft Learn)