3 September 2026LANTEK Computers
Break-fix is not a bad model. For a business of four people with three laptops and cloud email, paying somebody hourly when something breaks is the correct answer, and a monthly fee would be waste.
The problem is that businesses grow past it without noticing, because the model gives no signal. The invoice for reactive work looks the same at eight people as at thirty-eight, just more often, and nothing in it says you have outgrown the arrangement.
This is a method for working out where you actually are. No figures are given, because your rates and your staff costs are the only ones that matter.
The costs the invoice shows
Straightforward. Hours billed, plus call-out or travel, plus any minimum charge per visit, plus parts. Add up twelve months of it. That is the number most businesses have in mind when they say break-fix is cheaper.
The costs the invoice does not show
This is where the comparison actually lives, and all of it is calculable from things you already know.
Staff time waiting
The largest hidden item, usually by a wide margin. When something breaks under break-fix, the sequence is: somebody notices, tells somebody, that person telephones the provider, the provider schedules it, and the work happens. Meanwhile the affected person is either idle or working around it.
To put a number on it: take your last twelve months of support invoices and count the incidents. For each, estimate how many people were unable to work properly and for how many hours. Multiply by a rough loaded hourly cost for those staff.
Most businesses doing this exercise for the first time find this number is a multiple of what they paid the provider. It is not on any invoice, which is exactly why it is invisible.
The problems that are never reported
Under break-fix, reporting something costs money, so people do not report the small things. The machine that takes four minutes to log in. The printer that works on the third attempt. The share that has to be reconnected each morning.
Each of those is a few minutes a day. Multiplied by the person, by the working days, by the number of people with a similar annoyance, it becomes a substantial figure that nobody is tracking because no ticket was ever raised.
Ask your staff what they have stopped bothering to report. The list is usually longer than management expects, and it is a direct readout of the model’s incentives.
The same fault, repeatedly
Break-fix pays for restoring service, not for finding causes. So the fault that recurs monthly gets fixed monthly, and nobody is paid to ask why.
Go through your invoices and mark anything that appears more than twice. That is the cost of not diagnosing, and it is billed at the same rate each time.
Deferred maintenance
Patching, firmware, disk space, certificate renewals, backup verification. None of these generate a call while they are being neglected. They generate one large call later.
The pattern is consistent: nothing, nothing, nothing, then a failure that takes days and costs more than years of maintenance would have.
The security exposure
The one that does not average out. Nobody under break-fix is being paid to notice that a machine has stopped reporting, that multi-factor authentication was never turned on, or that a leaver’s account still works.
This does not show up as a cost most years. When it shows up, it shows up as the week described in a ransomware week, and no amount of hourly billing recovers that.
Unpredictability
Harder to quantify and real. A cost that arrives without warning in a month you had other plans for is worse than the same annual amount spread evenly, and it is why a bad month under break-fix tends to trigger the decision to change models.
The exercise, in one paragraph
Twelve months of support invoices. Count the incidents. For each, estimate people affected and hours lost, and cost that at your own staff rates. Add the annoyances nobody reports, at a few minutes each per day. Mark the recurring faults. Then compare the total against a managed quote applied to your real headcount and device count, per how to read an IT quote.
An afternoon’s work, and it produces a defensible number rather than an impression.
When break-fix is still right
It genuinely is, in these cases, and a provider who tells you otherwise is selling rather than advising.
Very few staff and very few systems. Everything already in cloud services with no server and no line of business application. Somebody in-house who is genuinely capable and has the time. Or a business where a day of downtime is an inconvenience rather than a cost, which does exist.
When you have clearly outgrown it
The signals are consistent, and none of them are about size alone.
You cannot say how old your machines are or what is under warranty. Nobody has restored from your backup. Staff have a mental list of things they no longer bother reporting. The same fault keeps recurring. You are surprised by support invoices. A day of downtime would cost real money. And nobody could tell you whether multi-factor authentication is on for everybody.
Three or more of those, and the arrangement is behind the business.
The middle option people forget
It is not binary. You can put the preventative work on a fixed fee, which is where the value is, and keep project work hourly. Monitoring, patching, backup verification, access reviews and an asset register on a monthly basis; migrations and installations quoted separately.
For a lot of businesses in the middle this is the honest answer, and it is worth asking for explicitly rather than accepting whichever package is on the sheet. If you have somebody capable in-house already, co-managed IT is the version of this that keeps them.
The point
Break-fix is not cheaper or more expensive in the abstract. It moves cost off the invoice and onto your staff’s time and your risk exposure, where nobody is counting it.
The decision is not about which model is better. It is about whether the costs you cannot see have grown past the ones you can.
What a managed service should cover is in what a managed IT service should include, and our plans are on Service Plans. If you want help doing the calculation on your own invoices, we will go through them with you, including where the answer is that you are fine as you are.