For an office of 10 or more people that depends on email, shared files and a line-of-business app all day, managed IT usually costs less over a real year, because the expensive events stop happening. For a very small shop with simple systems and real tolerance for downtime, break-fix is cheaper. The honest answer is arithmetic, so here it is.
The two models, in plain terms
Break-fix means you pay an hourly rate when something breaks, and nothing happens between calls. Managed IT means you pay a flat monthly fee per person, and in exchange somebody patches, monitors, backs up and answers the phone whether or not anything is on fire. One bills you for problems. The other bills you for the absence of problems. Everything else about the comparison follows from that difference.
The math, using numbers you can check
Most comparisons of these two models use invented numbers. We publish our prices, so this one can use real ones: our business packages run $65, $80 and $99 per seat per month, and our published hourly rates are $199 for remote business support and $219 on-site, billed in 30 minute increments, per the rate card. These are our rates, not market averages; swap in any provider’s numbers and the shape of the result holds.
Take a hypothetical 10 person office. On the middle package it pays $80 per seat, so $800 a month, $9,600 a year, and that fee includes the security stack: endpoint protection with EDR, email filtering, Microsoft 365 backup, identity protection and around-the-clock security monitoring. Now run the same office on break-fix across three kinds of year.
| Scenario | What happens | Break-fix spend | Managed spend |
|---|---|---|---|
| Quiet year | One remote hour a month, antivirus licenses for 10 machines at our published $95 per device per year | About $3,300 | $9,600 |
| Normal year | Two remote hours a month, four two-hour on-site visits, two machine setups with data migration, antivirus licenses | About $7,900 | $9,600 |
| Bad year | Everything in the normal year, plus one serious incident consuming about 30 on-site hours, plus recovery work | Over $14,000, before counting lost work time | $9,600 |
Read the table honestly. Break-fix wins the quiet year by thousands of dollars, and it can even win a normal year on raw labor. Managed IT wins the bad year, and the bad year is the one that decides whether the business has a bad decade. One serious incident, a failed server, a ransomware event, a dead drive with the only copy of the books, erases several quiet years of savings at once, and that figure still excludes the biggest line: a week of staff paid to not work. Data recovery alone starts at $199, and that is the floor, not the bill.
The deeper point is that the two columns are not buying the same thing. The break-fix column buys hours. The managed column buys hours plus the stack, the patching, the backup tests and the monitoring that make the bad year less likely to arrive at all. That is also the honest limit of the math: nobody can tell you in advance how many bad years you will have. What the models change is the odds.
Three questions that decide it faster than any spreadsheet
What does a full day of downtime cost you? Loaded payroll for everyone who cannot work, plus whatever revenue does not arrive. If that number is bigger than a month of managed service, the quiet-year savings are an illusion you are renting.
Who notices a problem first, you or the provider? Under break-fix the answer is always you, at the worst time. Under working managed service, a failing drive is a ticket on Tuesday instead of an outage on payroll day. If you are paying a monthly fee and still noticing problems first, you have the managed price with the break-fix experience, which is the worst seat in the house.
When did anyone last prove the backup restores? Not checked the dashboard. Restored a real file and opened it. If no one can name the date, you are running break-fix on the one system where finding out late is fatal, whatever your support model says.
Not sure which side of the math you are on? Book a quick scoping call and we will run your numbers with you, even if the answer is staying break-fix.
Where break-fix quietly fails
Nothing happens between calls. Patching, updates and account cleanup only occur when someone is paying by the hour for them, which in practice means they do not occur. Machines drift out of date until the drift becomes the incident. The clearest current example: computers still running Windows 10, which Microsoft retired from support in October 2025, keep working fine right up until the unpatched hole in them gets used.
Backups are assumed, not tested. Under break-fix, the first real test of the backup is the day you need it. A backup that has never been restored is a hope, not a plan.
Problems are found late. With no monitoring, the failing drive, the full disk and the dying fan all announce themselves at the worst possible moment, as outages instead of alerts.
The incentives point the wrong way. An hourly provider is paid more when your problems last longer. Most are honest anyway. But you are betting on character against economics, and over years, economics is patient.
The meter changes your behavior. When every call costs money, owners wait. Small problems get lived with until they compound into big ones, and the helpdesk call that would have been 20 minutes in March becomes a rebuild in August.
Where managed IT quietly fails
Fair is fair, and this is the part most MSP articles skip.
Paying for shelfware. Seats for people who left, tiers with inclusions nobody uses. Audit the seat count quarterly and insist the tier matches reality.
The contract outlives the quality. Service that was excellent in year one can coast by year three while the agreement auto-renews. Prefer short terms, and judge a provider by whether they will let the work speak instead of the contract.
Unlimited, with exclusions. Some agreements advertise unlimited support, then exclude projects, on-site time or anything interesting. Read the exclusions list before the inclusions list.
Automation without answers. A provider can monitor everything and still be impossible to reach. Monitoring is table stakes; the question is who picks up the phone, and how fast tickets actually close. Ask to see real numbers, and ask who you call when it matters.
None of these failure modes mean the model is wrong. They mean the contract and the provider need the same scrutiny the hourly invoice gets. A flat fee is only a good deal while someone is earning it.
Who should genuinely stay break-fix
Some businesses are correctly on break-fix, and an honest provider says so. Stay hourly if most of this is true:
- Five or fewer people, and no server, just laptops and cloud apps
- A full day of downtime is an annoyance, not lost revenue
- No regulator, insurer or client contract requires security evidence
- Somebody competent and reachable already handles your occasional breakage
- You verify your own backups by actually restoring a file now and then
- Cash flow genuinely cannot absorb a fixed monthly fee yet
If that is you, hourly is the right call, and we do that work at the published rates without trying to upsell you into a plan you do not need.
Staying break-fix well takes three habits, because you are now your own maintenance department. Keep a one-page inventory of every machine and account, so an incident starts with a list instead of a search. Put updates on a monthly calendar reminder and actually run them. And restore one real file from backup every quarter. Those three habits cost nothing and close most of the gap between the models for a genuinely small, simple office.
Who should not
Move to managed support when any of these is true: missing a payroll or shipping deadline costs real money; you hold regulated or client-confidential data; 10 or more people depend on the same systems; a cyber insurance questionnaire is asking about security controls like MFA, EDR and backup testing and nobody is sure of the answers; or the entire setup lives in one person’s head. Any one of these means the bad-year column in the table above is not hypothetical for you, it is scheduled.
The decision checklist
Score one point for each yes.
- Would a full day of downtime cost more than a month of managed service?
- Do 10 or more people depend on the same systems to work?
- Does anyone outside the business (regulator, insurer, client) require security evidence?
- Has it been more than a year since someone test-restored a backup?
- Is there a server, or an application that only runs on one aging machine?
- Did your last IT incident take more than a day to fully resolve?
- Does one person hold all the passwords and knowledge?
- Are there machines nobody has updated in months?
Zero to two: stay break-fix with a clear conscience, and test your backups. Three to four: get a managed quote and compare it against your last 12 months of IT spending plus downtime. Five or more: the quiet-year savings you are protecting do not exist; you are already paying managed prices in risk.
What this comparison does not claim
It does not claim managed IT is always cheaper; the table above shows the opposite for quiet years. It does not use market averages; the math uses our published rates because those are the ones we can print and you can check. And it cannot know your incident count in advance, which is the one variable that settles the question. We sell managed IT, so we have a side here. The checklist is how you keep our interest from deciding for you.
Talk it through with someone who does the work
We have run both models for Central New York businesses since 2005, rated 4.9 from 154 public Google reviews. Bring your last year of IT invoices to a short call and we will do the table above with your real numbers, in writing, before anything is signed.
Book a quick scoping call or call (315) 682-6372.
Questions owners actually ask about the two models
What does break-fix actually mean?
You call when something breaks, a technician fixes it, and you pay an hourly rate for the time. There is no monthly fee, and also no patching, monitoring, backup testing or account cleanup between calls. The name is literal: things break, then they get fixed, and the gap between those two events is yours to carry.
Can we mix the two models?
Yes, and many businesses should. Common mixes: a managed plan for the office plus hourly for projects, or keeping an in-house IT person while a provider supplies the overnight monitoring, patching and backup testing around them, which is usually called co-managed IT. The mix only works when the contract states plainly which side owns what.
Why do some IT companies refuse break-fix work?
Because unmanaged environments are where the worst emergencies live, and some providers only staff planned, monitored work. We take hourly work at published rates. But it is worth knowing that when a serious incident hits an unmanaged network, every provider in town will be triaging it behind their contracted clients.
How do we switch from break-fix to managed without a gap?
A reasonable provider starts with an onboarding pass: inventory every machine and account, document the network, verify the backups, and fix the dangerous stuff first. Keep your current arrangement until that pass is done. Get the onboarding scope, its cost and its end state in writing, and make sure the admin passwords end up owned by you, not by either provider.
