IT Support & Operations · 18 September 2026 · 9 min read

The IT Support Specialist's Guide to IT Administration and Microsoft 365

Yes, we have tried turning it off and on again. That was step zero.

There is a persistent myth that an IT support specialist is a person who reinstalls printer drivers and asks whether the cable is plugged in. I have done both of those things this decade, and I regret nothing — but it is a bit like describing a pilot as someone who operates a seatbelt sign.

Modern IT support sits on top of a full cloud platform. The same person who walks a user through a sign-in prompt is often the one administering Entra ID, enforcing compliance in Intune, tracing mail flow in Exchange Online, tidying SharePoint permissions, and writing the PowerShell that makes next month's version of the problem disappear entirely. This is what that job actually looks like — and how to do it well.

What an IT support specialist really does in 2026

The honest version of the job description has two halves. The first half is reactive: someone cannot work, and you make that stop being true. The second half — the half that decides whether your week is calm or chaotic — is administration: configuring the platform so the first half happens less often.

Early in my career I judged a good day by the ticket count. Fifteen years later I judge it by how quiet Monday morning is. That shift, from closing tickets to removing causes, is the entire difference between support as a helpdesk and support as an engineering discipline. I have written more about the mindset in how the IT support role is evolving.

The Microsoft 365 stack, from a support engineer's chair

In a Microsoft-based organisation, almost no incident stays inside one product. A "Teams problem" is frequently an identity problem wearing a headset. Here is how the pieces connect in day-to-day work:

Troubleshooting: read the logs before you trust the story

Users describe symptoms, not causes, and that is entirely reasonable — it is not their job to know that their laptop fell out of compliance at 03:14. A practical sequence: reproduce it, check Entra ID sign-in logs, check Intune compliance state, check the service health dashboard, then check the network path. Only after that should anyone recreate an Outlook profile, which is the IT equivalent of turning the sofa upside down to find the remote.

Identity first, always

Conditional Access is the single most powerful and most dangerous tool in the tenant. It can stop an attacker and, configured with enough confidence on a Friday afternoon, an entire sales department. Design it deliberately, test in report-only mode, and keep your break-glass accounts excluded and verified. My practical Conditional Access guide walks through the rollout, and Entra ID for small and medium businesses covers the smaller-tenant version.

Endpoints: managed, or merely hopeful

An unmanaged device is a promise that someone will behave. Intune turns that promise into a policy — compliance, encryption, patching, app deployment and Autopilot provisioning that lets a new starter unbox a laptop without a support visit. When I migrated 500 users to Intune, the technical work was the easy part; the sequencing and communication were where the project was actually won.

Security is not a separate task — it is how you do the task

Support work touches permissions constantly, which means support work is security work whether anyone writes that on the job description or not. Granting a shared mailbox, approving a sharing link, adding a guest, excluding a device from a policy — each of those is a small decision with a long tail.

The uncomfortable part is that most breaches do not need clever attackers, just patient ones and an environment where "temporary" access from 2021 is still active. This is exactly why Copilot exposes weak Microsoft 365 permissions so quickly: AI does not create new access, it just makes the existing access easy to find.

Automation, documentation and the art of not repeating yourself

PowerShell is the highest-return skill in this job. Bulk licence changes, offboarding, mailbox permission reports, stale device cleanup — all of it goes from an afternoon to a scheduled task. Automation does not replace engineers; it removes the work that was never engineering in the first place, a point I make in how AI is transforming IT support.

Documentation is the other half. Not a 40-page manual nobody opens — a short, findable note that says what broke, why, and what fixed it. Tribal knowledge is just a single point of failure with a pleasant name and a holiday allowance. When support demand grows beyond individual expertise, the next step is a documented technical support operating model and 90-day scaling plan.

User communication: the most underrated technical skill

You can resolve an incident perfectly and still leave the user convinced that IT ignored them. A single sentence — "we know what is wrong, here is the workaround, here is when it will be fixed" — changes the entire experience. Plain language, no blame, no acronym soup.

A practical IT administration checklist

None of these are glamorous. All of them prevent the incidents that ruin a quarter:

Practical takeaways

Frequently asked questions

Conclusion

The best IT support specialists are administrators who happen to be excellent with people — or people-focused engineers who happen to run a tenant properly. They fix the immediate problem, then quietly remove the conditions that created it. They automate the boring parts, document the useful parts, and explain both in language a finance director can follow.

That is the standard I hold myself to after 15+ years in IT, and it is the kind of team I want to build and be part of.

About the author — Davor Smajilovic, 15+ years of professional IT experience and 17 professional certifications including Microsoft, Cisco, ITIL and Skillsoft. Previously Technology Specialist at Lenovo (2022–2026), now open to new IT opportunities in Copenhagen. More on my skills or working together.