Managed IT Support

What “managed” is supposed to mean

Almost every IT company calls itself managed. Ask what the word means and the answer is usually about answering the phone quickly. Answering quickly is good, but it is not what managed describes. A responsive provider is still waiting for you to notice a problem and report it — by which point it has cost somebody a morning.

Managed means a defined set of things happen to your systems on a schedule, whether or not anyone rings. That machinery is what you pay for. Our plain-English guide to managed IT support covers the idea; this page describes the mechanics underneath it.

The machinery, piece by piece

Monitoring agents on your devices

A small piece of software sits on each laptop, desktop and server. It reports on the things that predict a failure rather than describe one after the fact: disk health warnings, free space trending steadily downwards, memory errors, machines rebooting unexpectedly, encryption switched off, a security service that stopped and never restarted, a backup job that has quietly failed four nights running.

The checks run continuously, and what they produce is a queue of signals rather than a room of people watching screens. That still changes the outcome: a failing drive announces itself in the logs days before it strands anybody, and a laptop six weeks behind on patching is visible long before it becomes the way something gets in.

Patching on a cycle rather than on a whim

Updates get applied to an agreed schedule instead of when a machine nags whoever is using it. That means a defined reboot window so nobody loses work mid-afternoon, a staged rollout so a bad update meets two machines rather than forty, the ability to hold back an update known to break something you depend on, and a view of which devices are current and which are not.

“We patch everything” is a claim. A list of the machines three months behind, with the reason for each, is a fact — usually a laptop belonging to someone who never reboots, or one sitting in a drawer since a member of staff left.

Alerts, and what an alert actually triggers

An alert is only useful if something happens next. Each type needs a response agreed in advance: this one raises a ticket automatically, this one is cleared by a script, this one gets a phone call because a server is about to stop. Without that, alerting becomes noise — an inbox of warnings nobody reads, which feels like cover while providing none. Tuning alerts to the ones that mean something is an early job on a new estate.

Tickets, and why a written history matters

Every request becomes a ticket, including those raised by monitoring rather than by a person. That is not bureaucracy for its own sake. A written history is the only way anyone notices that the same laptop has dropped off the Wi-Fi four times this quarter, or that three people reported the same slow application. Separately those are annoyances, fixed and forgotten. Together they are a fault with a root cause worth finding.

Asset inventory

An accurate list of what you own: every device, its age, specification, warranty status, who has it and what licences sit on it. That sounds administrative until you need it. It turns hardware budgeting into a replacement plan rather than guesswork, tells you which devices a departing employee holds, and stops you funding licences for people who left last year.

Backup verification, not just backups existing

These are different things, and the gap between them is where businesses get hurt. A backup that runs is a job with a green tick. A verified backup is one somebody has restored data from, having confirmed the files open. Backups fail silently for months — reporting success while writing to a full disk, or covering a server that no longer holds the data that matters. Managed support should include checking results and proving a restore periodically; how that is built lives on our backup and continuity page.

Onboarding and offboarding as a defined process

A new starter needs an account, a licence, the right file access, a configured device and the security policy applied — before their first morning, not during it. A leaver needs the reverse, promptly: sign-in blocked, sessions revoked, mailbox handled to your rules, device wiped, licence reclaimed.

Both are checklists, and the reason to run them as a defined process is that ad-hoc versions miss the same items every time — the shared mailbox nobody remembers, the app outside your main sign-in, the personal phone still collecting company email. The device build itself is covered by standardised deployment.

What the fixed fee changes about the relationship

This part is structural rather than promotional, and worth understanding before you compare quotes. Under pay-as-you-go, your problems are the provider’s revenue. Nobody sensible sets out to cause faults, but the incentive points the wrong way: a month where everything runs perfectly is a month with no invoice, and time spent preventing a fault is time that cannot be billed. A fixed monthly fee inverts that. Your problems become the provider’s cost. Every ticket we work is time we have already been paid for, so the cheapest possible month for us is one where your systems simply run.

That is why the unglamorous work — patching, verification, tuning alerts, documenting the estate — actually gets done. The model only holds if the fee is realistic, though. Priced too low, it creates pressure to do less — the same problem wearing a different hat. Be wary of a number that looks too good for the scope attached to it.

What managed IT support does not include

A provider who implies the monthly fee covers everything is setting up a difficult conversation later. The boundary usually sits in roughly the same place:

• Project work — migrations, server replacements, office moves and rollouts are quoted separately
• Hardware — laptops, servers, switches and firewalls are your assets; we specify, source and configure them, but the kit is bought
• Third-party licences — Microsoft 365, line-of-business software and security subscriptions are billed as licences, through us or direct
• Development work — changes inside your own bespoke software normally belong to the vendor who wrote it
• Out-of-hours cover — available, but agreed and priced deliberately rather than assumed

None of that makes the fee less valuable. It means you can read a proposal and know what the number covers.

Managed, outsourced, or alongside your own IT person

These get used interchangeably and they are not the same. Managed IT support is the ongoing service: the machinery above, plus a helpdesk your staff use directly. Outsourced IT is the whole department — that same service plus strategy, suppliers, procurement, budgets and a senior contact for the decisions. Managed keeps things running well; outsourced also decides what they should be. If you already employ someone technical, co-managed support puts a team behind them rather than replacing them. All four models are compared side by side on our business IT support page.

Getting started

1. A review of what exists — devices, accounts, network, licensing, security, and whether the backups would restore
2. A plain list of what we found — ranked by risk, with the urgent separated from what can wait
3. Onboarding — monitoring deployed, patching brought under control, documentation and the asset list built
4. Day-to-day support — your team has one place to go, and the scheduled work carries on behind it

If you are currently the person everyone asks when the Wi-Fi plays up, that is the job we take on. Get in touch and we start there.