Ten questions on whether you have outgrown calling someone when it breaks. Built for small businesses across Chester County PA, northern Delaware, and northern Maryland, though the questions apply anywhere.
10 questions, about ninety seconds. Everything runs in your browser and nothing you enter is sent anywhere.
Ordered by how much weight each one carries. Disagree with any of them and the recommendation changes, which is rather the point of showing you the reasoning.
Somewhere around five or six people, IT stops being occasional and becomes a steady trickle. The trickle lands on whoever is best with computers, and it is rarely their job.
Try this: Ask that person roughly how many hours a month they spend on other people’s computer problems. Multiply it by what an hour of their actual job is worth.
What changes if this is handled: Either fund the role properly or move the work outside. What does not work is leaving it where it is and hoping.
Twenty-plus people generate enough accounts, devices, joiners and leavers that the work is now continuous. Businesses this size usually have the volume without the structure, so everything happens reactively.
Try this: Count the IT things that happened last month: new starters, leavers, password resets, anything broken. If nobody can produce that list, the volume is not being tracked, which is itself the finding.
What changes if this is handled: A defined arrangement, whether that is internal, external, or both. The specific shape matters less than somebody owning it.
At this size the question is not whether IT needs managing but who is doing it and whether they have what they need. Businesses that reach this point without deciding usually have a very overworked internal person and no succession plan for them.
Try this: Ask what happens if the person who knows how everything works is off for two weeks. If the honest answer is "we wait", that is the risk.
What changes if this is handled: A real arrangement with documentation behind it, so the business does not depend on one person’s memory.
Past about ten, the list stops living reliably in someone’s head. Machines then age, fall behind on updates, or sit unused with live accounts on them, and none of it is visible until something goes wrong.
Try this: Try to write down every device the business owns, from memory, in five minutes. Then compare it to what is actually plugged in.
What changes if this is handled: An inventory first. It is dull and it is the thing everything else depends on, including knowing what you are paying for.
At this count, keeping machines updated by walking round is not realistic, so it stops happening. The gap between "we install updates" and "our machines are up to date" gets wide, quietly.
Try this: Ask for a list of your devices with the date each last updated. The time it takes to produce that answer tells you more than the answer does.
What changes if this is handled: Central monitoring and patching. This is the point where a managed arrangement starts saving money rather than costing it.
Not being able to answer is the answer. You cannot secure, budget for, or replace things you have not counted, and you are probably paying for some of them twice.
Try this: Compare your software and licence bills against your headcount. Unused licences for people who left are the usual first thing to fall out.
What changes if this is handled: A discovery exercise, then a maintained inventory. It usually pays for itself in cancelled licences before it does anything else.
Walking over to someone’s desk stops being the fix. Home working also puts company data on connections and machines nobody chose, which is a different problem from the one your office setup was built for.
Try this: Ask a home worker what happens when their machine will not start on a Monday morning. Count the hours in the answer.
What changes if this is handled: Remote support and remote management, so where someone sits stops being the deciding factor.
Every site has its own network, its own equipment, and its own set of things quietly failing. Whichever site the helpful person is not at gets the worst of it, and its problems get reported latest.
Try this: Ask people at your smallest site what they currently just live with. It is usually a longer list than at head office, and none of it has been reported.
What changes if this is handled: Remote monitoring so every site is visible from one place, rather than support quality depending on geography.
At this spread, hands-on support is no longer the model, and anything that depends on someone being physically present will not scale. It also becomes very hard to say what you own and where it is.
Try this: Pick your most remote site or a field worker and time how long a straightforward problem actually takes to resolve, start to finish.
What changes if this is handled: Remote-first management with on-site work as the exception, and equipment standardised so a replacement is a swap rather than a project.
Break/fix is not a bad arrangement, it is a specific one: you pay for repairs and nobody is paid to prevent them. That works while problems are rare. The incentives also point the wrong way, since the provider earns more when things break.
Try this: Add up last year’s IT invoices. Then ask how many of those visits were for something that could have been caught earlier.
What changes if this is handled: Not necessarily a change. If the total is low and the response is good, stay where you are. If it is climbing, a fixed monthly cost is worth pricing against it.
This is the most common arrangement I find and the least visible. It costs you their real work, it leaves nothing written down, and it ends the day they leave or finally refuse.
Try this: Ask them, privately, how they feel about it. The answer is usually more strongly worded than anyone expects.
What changes if this is handled: Give the work to someone whose job it is. That can still be part time, and it does not have to mean a full plan.
When nothing gets reported because reporting achieves nothing, people build workarounds instead. Those workarounds become how the business runs, and they are invisible until someone new arrives and asks why.
Try this: Ask your team what they have stopped bothering to report. Expect a longer list than you were expecting.
What changes if this is handled: One route in, and someone at the other end of it. This is the cheapest change on the list and usually the one people notice most.
Monthly is the point where these stop being events and start being a line item, except nobody is adding it up because it never arrives as a bill.
Try this: Use the downtime calculator with your real numbers and a monthly frequency. The annual figure is usually the part that surprises people.
What changes if this is handled: Worth measuring before deciding anything. If the annual figure is small, carry on. If it is not, you now have the number to compare against a quote.
Weekly problems are a symptom rather than a run of bad luck. Usually it is ageing hardware, a network that was never designed, or the same fault being re-fixed because nobody recorded it the first three times.
Try this: Ask whether the last five problems were five different things or the same thing five times. If nobody can say, nothing is being recorded.
What changes if this is handled: Find the underlying cause rather than the incidents. That needs a record, which is the thing reactive support never produces.
At this frequency people have stopped noticing, which is worse than complaining. The lost time is now baked into how long everything takes, and it will not show up anywhere you are looking.
Try this: Ask how long a routine task is expected to take, then ask someone who does it daily how long it really takes. The gap is often IT.
What changes if this is handled: This is usually a handful of specific causes rather than everything being bad. Finding them is a short piece of work and the results tend to be immediate.
This is tolerable for a broken printer and expensive for anything a person needs to do their job. The cost lands on you rather than on whoever is slow to arrive.
Try this: Multiply your typical wait by how often problems happen, and by what an hour of that person costs. That is the annual price of the current response time.
What changes if this is handled: An agreed response time, in writing, rather than an informal expectation. It is the difference between a promise and a habit.
Unpredictable is worse than slow. You cannot tell a client when you will be back, you cannot decide whether to send someone home, and you cannot judge whether the arrangement is working.
Try this: Look up when you reported your last three problems and when each was actually resolved. Three very different numbers is the finding.
What changes if this is handled: A written response commitment. Any provider unwilling to put one in writing has told you something useful.
Chasing is a reliable sign that requests are being held in someone’s head rather than in anything that tracks them. It works until that person is busy, which is exactly when you need them.
Try this: Ask what happens to a request when the person who took it is on holiday. If the answer involves their inbox, there is no system.
What changes if this is handled: Anything that tracks requests and reports back without being asked. The tracking matters more than the tool.
If nothing progresses unless you push it, then managing your IT provider has become part of your job. That is time you are already paying for, on top of the invoice.
Try this: Count the follow-up emails and calls you sent about IT last month. That is the unbilled hours you are contributing.
What changes if this is handled: A provider that reports without being asked, or an internal system that does. Either way you should not be the one holding the list.
Break/fix answers the phone. It does not plan a move, size a network for twice the headcount, or work out what a new system needs before you have signed for it. Those get handled at the last minute, which is where the cost is.
Try this: Ask who is going to work out the IT side of it, and when. If the answer is "we will sort it nearer the time", the budget for it does not exist yet.
What changes if this is handled: Get the IT requirements scoped before the commitment rather than after. That is advisory work and it is usually the cheapest part of the project.
Individually these are manageable. Together they interact: new people need devices and accounts, a new site needs a network, new software needs both. Handled separately and late, they collide.
Try this: Put the changes on one calendar with dates. Overlaps you had not noticed are usually visible within a minute.
What changes if this is handled: One plan covering all of it, sequenced. This is the clearest case on the list for having someone whose job includes looking ahead.
The application asks whether multi-factor sign-in is enforced and whether backups are tested. You sign to say yes. If an answer is wrong at claim time, the insurer can decline. Somebody has to keep those answers true all year, not just on the day you signed.
Try this: Read last year’s application again and pick the answer you are least sure of today. That is the one to check.
What changes if this is handled: Make each answer somebody’s responsibility with a date attached. The paid Network Security Assessment exists specifically to produce that document.
A security clause in a contract is a commitment you can be held to commercially. Reactive support does not evidence anything, so when a client asks for proof there is nothing to send, and that conversation tends to happen during a renewal.
Try this: Find the security clause in your largest contract and read it. Then ask who would answer it if the client asked tomorrow.
What changes if this is handled: Someone owning the requirements and keeping evidence as they go. Assembling it retrospectively costs several times more.
The obligation does not adjust for how your IT happens to be arranged. It stays yours regardless of who is or is not looking after it, and "nobody was managing it" has never been a defence.
Try this: Ask who would produce the evidence if it were requested this week, and how long it would take them.
What changes if this is handled: A defined arrangement with documentation. What that has to cover depends on your regulator, and it is worth reading their requirements rather than anyone’s summary of them, mine included.
Buying under pressure means paying whatever the day costs, taking whatever is in stock, and losing a person for however long setup takes. The same machine bought in a planned month costs less and interrupts nothing.
Try this: Work out what the last emergency replacement cost you all in: the machine, the lost day, and the setup time. Compare it to replacing two a year on a schedule.
What changes if this is handled: A simple refresh schedule spread over a few years. It turns a series of bad surprises into a budget line.
Without a plan, machines are replaced only after failing, so the average age of everything creeps up together. That means the failures arrive in a cluster rather than spread out.
Try this: Find the age of your five oldest machines. If they are all similar, they were bought together and they will fail together.
What changes if this is handled: Age the estate on paper, then spread the replacements. The first pass takes an afternoon and settles the budget conversation for years.
This usually means no inventory, which means no plan is possible yet. It also means hardware is probably still in service past the point of getting security updates, which quietly turns a budgeting question into a security one.
Try this: Check what version of Windows the oldest machines run and look up its support end date. That date is a deadline and it will not move.
What changes if this is handled: An inventory with ages and warranty dates. It is the smallest useful first step and everything else depends on it.
That is a real result and not a polite one. On what you have told me there is no case for changing anything. Worth running again if you take on people, add a site, or notice the problems starting to cluster.
I will go through your result, tell you honestly whether a plan is worth it for a business your size, and give you a real number rather than a range. If the answer is that you are fine as you are, I will say that instead.
Book a 15 minute callI will send your score and the full list of findings as a document you can keep, forward to your provider, or put in front of your insurer. One email, and I do not add you to a list.
It assumes you run the business rather than the network. If you can answer questions about how your office works, you can complete it.
For a small team in one office with a provider who answers the same day, paying by the hour is cheaper and simpler. This tool will tell you that if it is true. I would rather say so than sell you a plan you do not need, and you would find out either way.
Nobody wakes up having outgrown their IT arrangement. It happens one hire, one site and one system at a time, and the moment it stops working is usually the moment something goes badly wrong rather than the moment it stopped being enough.
When there is no plan, the cost turns up as chased tickets, someone in accounts fixing printers, and work that quietly does not get done. None of that appears on a statement, which is the only reason it feels free.
The result names the specific things pushing you one way or the other, so you can disagree with any of them. A recommendation you cannot interrogate is not worth much.
Every answer carries a weight based on how much continuous work it creates, because four people in one office and forty across three sites are not the same question with a different number in it.
| What is missing | Weight | Example from this check |
|---|---|---|
| High-risk control missing | 10 points | Nobody owns IT, problems most weeks, or several sites |
| Moderate gap | 5 points | An outside provider you call, or replacing hardware only when it fails |
A high score means you are fine as you are. This runs the same way round as the other checks: high means your current arrangement still covers you, low means the load has outgrown it. The top result is "break/fix still fits", and a small single-site business with a provider that answers the same day genuinely lands there.
Nothing here is weighted as critical, on purpose. On the health check, one missing control can cap the rating, because a missing backup really is categorical. Nothing here is. It is all a matter of degree, so no single answer can push you to the strongest recommendation on its own. That has to be earned by accumulating real load, which matters when the person publishing the tool sells one of the outcomes.
It weighs up your situation, not your provider. Plenty of businesses that look like they have outgrown break/fix are actually fine because their provider is excellent, and plenty that look comfortable are being quietly let down. It also cannot price anything: what a plan would cost depends on your users, devices and risk, and anyone quoting from ten multiple-choice answers is guessing.
No, and the top result is literally "break/fix still fits". A small team in one office with a provider that answers the same day scores in the eighties and gets told to stay where they are. The result also shows you every factor behind it, so you can disagree with any of them.
Break/fix means you pay for repairs when something goes wrong, and nobody is paid to prevent them. Managed means a fixed monthly fee covering monitoring, maintenance, and support, so the incentive moves to keeping things working. Neither is better in the abstract. Which one is cheaper depends entirely on how often things go wrong and how much it costs you when they do.
Each answer carries a weight based on how much continuous work it creates. A higher score means your current arrangement still covers you. A lower score means the load has outgrown it. Nothing here is weighted as critical, deliberately, so no single answer can push you to the strongest recommendation on its own.
No, and be wary of anything that claims it can from ten multiple-choice answers. Real pricing depends on your user count, device count, servers, sites, and what you actually need covered. The plans page shows what is included, and a fifteen minute call gets you a real number.
The parts that genuinely need to be continuous, usually monitoring, patching, and backup, handled on a plan, with everything else staying hourly. It suits businesses that have outgrown pure break/fix but do not need a full outsourced IT function. It is the most common honest answer for a business between about eight and twenty people.
No. The whole result appears immediately and nothing you enter leaves your browser. Email is only for the written version, which is optional.
Those are genuinely different problems and it is worth separating them. This tool weighs the arrangement. The page on switching providers covers the signs that the problem is who you are with rather than how you are set up, and it gives you a test to run for each one.
All free, all a couple of minutes, and none of them ask for anything before showing you the answer.
Twelve questions on the tenant settings that decide whether one stolen password becomes an incident.
Start itTwelve questions on the difference between having backups and being able to use them.
Start itBuild a checklist for a departure in the order the steps actually need doing.
Start itI will walk through what came back, tell you which items genuinely need attention and which can wait, and you are under no obligation after it.
Free, no obligation, and you talk directly with a senior engineer.