What managed IT services actually cost — and what the per-user number hides

Managed IT is typically quoted at $100–$400 per user per month. That range is wide enough to be nearly useless on its own, because two providers quoting the same number can be selling substantially different things. What matters is what sits inside the number — and what has been left outside it.

The ranges, by tier

Small organisations pay more per user. Fixed costs — onboarding, documentation, tooling minimums — spread across fewer people. Below roughly fifteen users most providers apply a monthly minimum regardless of headcount.

The five things quoted separately

This is where proposals diverge most, and where the comparison usually goes wrong:

1. Onboarding. A one-off charge covering documentation, agent deployment and remediation of what they inherit. Often $50–$150 per user, occasionally waived on longer terms. A provider quoting zero onboarding is either absorbing it into a higher monthly rate or not doing it properly.
2. Projects. Migrations, office moves, hardware refreshes are almost never in the monthly fee. Ask for the project day rate up front — it varies far more between providers than the per-user figure does.
3. After-hours and onsite. "24/7 support" sometimes means 24/7 monitoring with business-hours response. These are very different products. Get the answer in writing.
4. Licences. Microsoft 365, endpoint protection and backup licensing may be bundled or billed at cost plus margin. Bundled is not automatically cheaper; it is simply harder to compare.
5. Security tooling. MDR, SIEM and phishing simulation are increasingly sold as add-ons rather than included, even in tiers marketed as security-led.

Comparing two proposals fairly

Reduce both to the same three-year figure. Nothing else makes them comparable:

(monthly per user × users × 36)
  + onboarding
  + (expected project days × day rate)
  + annual licence cost × 3
  = three-year total

Then ask each provider the same five questions. The answers separate them faster than the pricing does:

1. What is your guaranteed response time, and what happens if you miss it? An SLA with no remedy is a statement of intent.
2. Who owns the documentation and admin credentials? If the answer is not "you do", changing provider later will be expensive.
3. What is the exit process? Notice period, offboarding cost, data handover format. Ask before signing, not during a dispute.
4. How many users per engineer? Anything above roughly 150:1 tends to show up as slow response.
5. When did you last restore a client's backup, and how long did it take? Testing restores separates providers who monitor backups from providers who only tick that they exist.

Against the in-house comparison

The honest version of a calculation frequently presented dishonestly. A single systems administrator in the US runs about $96,800 median salary, and $130,000–$150,000 fully loaded once benefits, tax, tooling, training and recruitment are counted.

For a 25-person business, managed IT at $138 per user is roughly $41,400 a year against $130,000+ for one administrator. The MSP case looks overwhelming — but the comparison is not like for like, and a provider presenting only this half is selling, not advising.

What the comparison omits:

• An internal hire gives you dedicated, context-rich attention. An MSP gives you shared attention across a broader skill set.
• The economics invert with scale. At 150 users, $130 per user is around $234,000 a year — more than two internal engineers.
• Most organisations above roughly 75 users end up co-managed: internal staff for day-to-day and institutional knowledge, MSP for depth, out-of-hours and specialisms.

Under about 50 users the MSP case is usually genuine. Between 50 and 100 it depends on how much internal knowledge you have. Above that, do the arithmetic properly before accepting either answer.

Why the cheapest quote is usually the most expensive

A quote materially below the others is not generosity. It is one of:

• Higher user-to-engineer ratios, which surface as response times weeks after you sign.
• Break/fix in disguise — a low monthly fee where most real work is billed hourly.
• Thin security, with MDR, awareness training and tested backups all as extras.
• Junior-only first line, so every non-trivial ticket queues for escalation.

The reasonable move is not to pay the most. It is to normalise every quote to the same scope and compare on that.

If you are early in this

You do not need a decision immediately — you need a scope. Write down your user count, what must be covered outside business hours, any compliance obligations, and which systems genuinely cannot go down. Send that same document to every provider you approach. Quotes written against one scope are comparable; quotes written against three different assumptions are not.

Happy to be one of the providers you send it to — tell us what you are working with and we will quote against your scope rather than a template.