Choosing an IT support provider is one of those decisions that's easy to get wrong and hard to fix. The contract is typically 12-24 months, the switching cost is real, and the pain of a poor provider only becomes apparent gradually — one slow response, one missed backup, one poorly handled incident at a time.
These are the questions I ask when I'm evaluating providers on behalf of clients. They won't guarantee a good outcome, but they surface the issues that matter.
1. How Many Clients Does Each Engineer Handle?
This is the most direct indicator of whether a provider can give you attention. A managed service provider with 50 engineers and 5,000 client companies is running at 100 clients per engineer. That's a volume business where your 30-person firm is not going to get a lot of proactive attention.
There's no universal right number, but I'd be sceptical of anything above 80 active clients per engineer for a service that promises proactive management. Below 40-50 is genuinely good.
2. Who Will Actually Handle Our Day-to-Day Issues?
Many providers sell on the strength of their senior engineers and then allocate day-to-day tickets to junior staff or offshored helpdesk. That's not necessarily bad, but you should know what you're buying. Ask specifically: when a staff member calls with a problem, who answers? What's their qualification level and experience? When do issues get escalated, and to whom?
3. What Are Your Actual Response and Resolution Times?
Response time means someone acknowledges your ticket. Resolution time means the problem is fixed. They're different numbers and both matter.
Ask for the SLA in writing, for different severity levels. Then ask what their actual performance against those SLAs was in the last quarter. A provider confident in their performance will share this. One who hedges or doesn't have the data is telling you something.
4. Can We Speak to Three Current Clients of Similar Size?
References are standard. Three specific clients similar to you in size and sector, who you can actually call and ask direct questions, are meaningful. Testimonials on a website are not.
When you speak to references, ask: have there been any significant incidents in the past year? How were they handled? Is there anything you'd change about the service?
5. What Does Onboarding Look Like?
A provider who just shows up with a remote management agent and calls it onboarding isn't going to understand your systems. Good onboarding involves documenting your infrastructure, understanding your critical systems and processes, meeting key staff, and establishing a baseline for what normal looks like.
Ask how long onboarding takes and what it involves. A week isn't long enough to do this properly for most businesses.
6. How Do You Handle After-Hours Emergencies?
If your server goes down at 7pm on a Friday, what happens? Some providers offer 24/7 coverage; others are office hours only with emergency-only out-of-hours cover; others don't cover outside business hours at all.
Match this to your actual risk exposure. A business that can tolerate IT failure overnight is in a different position to one running customer-facing systems around the clock.
7. Who Owns Our Data and Documentation if We Leave?
This question makes providers uncomfortable, which is exactly why you should ask it. All your system documentation, passwords, network diagrams and configuration information should be yours. Some providers deliberately make this proprietary to create switching friction.
Get the answer in writing in the contract.
8. What Is the Notice Period and Termination Clause?
Read the termination clause carefully before signing. Some contracts have rolling notice periods, minimum contract terms, and early termination fees that make switching expensive even if the service is poor. A good provider is confident enough in their service to offer reasonable exit terms.