Most companies pick a managed IT provider the way they pick a roofer. Three quotes, a gut feeling about who seemed most competent on the phone, and a decision made on price because the proposals were too different to compare any other way.
That works badly here, because the thing you're actually buying is a promise about what happens on your worst day, and every proposal in front of you claims the same thing. Proactive monitoring. Enterprise-grade security. Rapid response. The words are identical across all three.
These are the ten questions that separate them. They're ordered roughly by how much money a wrong answer costs you, and each one includes what a good answer sounds like — because a provider can technically answer "yes" to most of these while committing to nothing.
Bring this to the meeting. Ask for the answers in writing.
Almost every provider will tell you they respond quickly. Far fewer will put a number in the agreement, and fewer still will define what the number measures.
Ask which document the commitment lives in, how priority levels are defined, and what happens when a target is missed. "Response" often means an automated ticket acknowledgment, not a human working the problem. Those are different things and only one of them helps you.
A good answer names a document, distinguishes response from resolution, and defines priorities by business impact rather than by technical category. A production line down and a single user's mailbox should not be the same priority.
You are buying a team, and you should know its shape. Is there a named account owner? Are you routed to whoever is free, or to people who have seen your environment before? Is any part of support offshore or subcontracted, and if so, which part?
None of those answers is automatically disqualifying. An offshore first tier can be perfectly good. What matters is that you know, in advance, rather than discovering it during an outage.
A good answer is specific and volunteered. A provider who deflects this question is telling you something.
Two proposals can look identical and differ enormously here, and it matters far more in a plant or a warehouse than in an office.
A failed switch in a production cabinet, an access point that needs relocating above racking, a cable run damaged by a lift, a new line that needs drops pulled — none of that gets solved over a remote session. If the agreement is silent on on-site work, assume it's billable, and ask for the hourly rate and the travel policy before you sign rather than after.
A good answer states plainly whether visits are included, capped at a number per month or quarter, or billed, and gives the rate. Ask how far away the nearest technician actually sits. We work through how to size your own on-site need here.
"Enterprise-grade security" is not an answer. Endpoint detection and response, a security operations center and who staffs it, multi-factor authentication, email filtering, DNS filtering, patch management, backup with tested restores, and user awareness training are all separable line items, and providers bundle them very differently.
Two quotes for the same company can cover materially different amounts of security at the same monthly price. Get the list, then compare the lists rather than the totals.
A good answer is an itemized list you could hand to your insurer. That's not incidental — cyber insurance renewals increasingly ask for exactly this inventory.
Ask specifically about structured cabling, wireless site surveys, access point placement, and network refresh work.
Many providers manage the switch and hand off everything physical to another company. That seam doesn't show up in a proposal — it shows up in a scheduling delay, when the firm you need next has its own queue and no agreement with you. In a facility with high-bay racking, moving equipment, or washdown areas, that's most of the hard problems.
A good answer is either "we do it ourselves, here's the crew" or a straight acknowledgment of who the partner is and how escalation works. Vagueness here usually means a subcontractor.
This question sorts providers faster than any other, because it has no good generic answer.
Every plant has them. Machine-attached PCs running an operating system the equipment vendor validated years ago. HMIs that will not survive a reboot in the middle of a run. An embedded controller whose support contract voids if you touch the software. Cloud and hybrid environments raise a version of the same problem from the other direction.
A provider who says they patch everything on a standard cycle either hasn't worked in your kind of building or is about to cost you a shift.
A good answer describes segmentation and compensating controls rather than universal patching, distinguishes IT systems from OT systems, and asks you questions back about your change windows.
The first sixty days should produce an asset inventory, a network diagram, a list of what is unsupported or past end of life, and documented escalation paths. That documentation is worth real money and you should own it outright.
Ask directly: if we part ways in two years, what leaves with us? Credentials, diagrams, and inventories should be yours. Some providers treat their documentation platform as proprietary and hand you very little.
A good answer commits to specific deliverables on a timeline and confirms you own the output.
Managed IT that only closes tickets is a help desk with better branding. The part that changes your budget is someone who tracks hardware and software lifecycle, watches what is aging out, and brings you a picture before a failure forces the decision.
Ask what that role is called, who fills it, how often you meet, and what you walk away with. Ask whether the person in the room can talk about your business or only about your servers.
A good answer describes a recurring review with an agenda and an output document. Research published by the Global Technology Industry Association in July 2025, based on a survey of 720 small and mid-sized businesses, found that 80% said they have some or a lot of room to improve on technology vision and strategy — which is precisely the gap this function is supposed to close.
Read the renewal clause before the pricing. Look for auto-renewal windows that require notice months in advance, early termination fees, and whether rates can change mid-term.
Then ask the offboarding question directly: what does a transition to another provider look like, what do you hand over, and is there a fee for it. A provider confident in their work will answer this without tensing up.
A good answer treats the exit as a normal part of the agreement rather than an insult.
Two halves, and both matter.
Like ours: a provider who has supported manufacturers understands production networks, change windows, and why the plant manager gets a say in the maintenance schedule. One who has only supported professional offices will learn on your equipment. What actually changes between a plant, a clinic, and a distribution center is here.
Near us: ask where their technicians are based and what a dispatch to your address actually looks like. Distance isn't automatically disqualifying — it's just a fact you want in hand before an emergency.
Ask for references you can call from companies of similar size and industry, and ask what certifications the firm holds and which individuals hold them.
A good answer produces named references and specific credentials rather than logos on a slide.
Ask whether the provider is profitable.
It sounds rude, and it isn't. ConnectWise's Service Leadership Index reported that in Q4 2024, average adjusted EBITDA across managed service providers was 11.1% — down from 12.2% the prior quarter — and 18% of providers were operating at a loss. A provider priced well below the others is either running a narrower scope than you understood or running unprofitably, and both of those become your problem in year three, when your entire environment depends on them still existing.
You aren't asking for their books. You're asking whether the number in front of you is sustainable.
Normalize everything to total monthly cost, then line up the answers to questions 3, 4, and 9 — on-site coverage, security stack, and contract terms. In our experience that's where most of the difference between two quotes turns out to sit, far more than headcount or device counts do.
If a provider will give you a monthly number without assessing your environment first, they are guessing, and you will meet the real number later as a change order.
What should I look for in a managed IT provider?
Written response commitments, a clear answer on whether on-site support is included, an itemized security stack, defined handling for systems that can't be patched normally, and a contract you can exit. Price is the least useful comparison point because scopes differ so widely.
Should price be the deciding factor?
No, because two monthly numbers built on different scopes aren't comparable. For reference, in Kaseya's 2026 State of the MSP survey of 1,061 managed service providers, 48% reported that their typical small and mid-sized client contract falls between $1,000 and $5,000 per month — but where a given company lands depends far more on security scope and coverage hours than on size. We break the benchmark data down here.
Should I choose a local IT provider or a national one?
It depends on how often your problems are physical. Offices with cloud-based applications can be supported well from anywhere. Plants, warehouses, and multi-building sites need someone who can be on the floor, which makes local presence a practical requirement rather than a preference.
How long does it take to switch managed IT providers?
Typically weeks rather than months: discovery and documentation, tooling deployment, a transition period with the outgoing arrangement still available, then steady state. The main thing to secure is that your documentation and credentials transfer to you.
What questions reveal the most about a provider?
Questions 6 and 9 — how they handle systems that can't be patched on a schedule, and what happens when you leave. Neither has a generic answer, so both are difficult to bluff.
CTC Technologies supports Michigan businesses from an Ann Arbor headquarters, with technicians who drive to your building. We manage networks and we also build them, so the cabling, the wireless survey, and the network refresh come from the same team that answers the ticket.
Start with an assessment rather than a proposal. We'll inventory what you have, document how it connects, and tell you where the risk sits. You keep the documentation either way.
Request an IT assessment or call 734-408-0200. Managed IT across Michigan, including Detroit, Troy, Warren, Livonia, Sterling Heights, and Ann Arbor.
