IT outsourcing sounds simple: you pay someone to look after the computers and stop worrying about it. In practice, if you do not know what to look for, you can end up with a company that sells brilliantly and delivers poorly. Here are five specific traps and how to avoid them before you sign anything.
Trap 1 — “We’ll get back to you as soon as we can”
The problem
the IT partner has no defined response time. “As soon as we can” means “when it suits us”. At 4:30 pm on a Friday, with the server down, it turns out that “as soon as we can” means Monday morning.
How it plays out
the company calls, leaves a message, waits. Nobody calls back. An hour later it calls again. An engineer turns up in the evening, or the next day. Cost of the downtime: thousands. The partner’s liability: none, because nothing was in the contract.
What to check in the contract
- time to first contact (for example 15 minutes during working hours, 30 outside)
- time to start work, not just to “get in touch”
- the definition of a critical incident versus a standard one
- penalties for breaching the SLA
If the partner will not put concrete numbers in the contract, that is an answer in itself.
Trap 2 — An invoice that makes no sense
The problem
you get an invoice for an amount with no explanation of what it covers. Or worse, a detailed breakdown along the lines of “labour: 4h × PLN 150” with no description of what was actually done.
How it plays out
you do not know what you are paying for. You do not know whether it was necessary. You cannot judge whether the price matches the work. You cannot challenge mistakes without full documentation.
What to check
- whether the partner provides a report of work done: what, when, on which device
- whether the scope of the monthly plan is clearly defined in the contract
- whether there is a ticketing system you can access as a client
A decent IT partner should be able to show you the history of every ticket, the response time and a description of the work. That is not a luxury, it is the basis of accountability.
Trap 3 — Firefighting instead of prevention
The problem
the partner reacts to failures but never does anything to stop them happening. The server goes down once a quarter and every time there is a “repair”. Nobody asks why it keeps going down.
How it plays out
downtime costs accumulate. Hardware ages with no replacement plan. Software is not updated because “it works”. The first serious failure or ransomware attack reveals that nothing was done for years beyond putting out fires.
What to check
- whether the partner proposes regular infrastructure reviews, quarterly for instance
- whether you get a report with recommendations after each review
- whether the partner keeps an asset record with purchase dates and planned replacement
- whether software updates are part of the plan or charged extra
Ask outright: “What does proactive care of our infrastructure look like at your end?” A vague answer is the signal.
Trap 4 — Paying by the hour with no cap
The problem
hourly billing with no fixed monthly plan. Every call, visit and email produces a new invoice. In a bad month you pay three times what you paid in a good one, with no way to plan a budget.
How it plays out
staff are afraid to call IT because “every ticket costs money”. Problems get reported too late, or not at all. Small matters grow into serious failures. And when the IT company is stretched across too many clients, your “cheap” ticket lands at the back of the queue.
What to check
- whether the offer includes a fixed monthly plan with a defined scope
- what is inside the plan and what is charged extra, new hardware or complex rollouts for example
- what the maximum cost of additional work in a month can be
- whether the hourly rate is agreed up front and in writing
A monthly plan is not just convenience. It aligns the partner’s incentives with yours: the fewer failures, the less work at the same fee.
Trap 5 — A different person every time
The problem
you call and reach a different engineer each time, one who does not know your company’s history. You explain the network layout, the quirks of your systems and the earlier problems from scratch. You lose time and the partner loses context.
How it plays out
an engineer without ticket history treats the symptom, not the cause. The same problem returns every few months. Nobody sees the pattern, because each person saw only one incident. The network grows chaotically, because every engineer “fixed it their own way”.
What to check
- whether the IT company runs a ticketing system with the full history of each client
- whether infrastructure documentation is produced during onboarding
- whether you, as the client, can look into the ticket system and see the history
An IT company serving dozens of clients without a proper ticketing system is a company running on intuition. Intuition fails on a Friday evening.
How to assess a partner before you sign
Before you put your name to anything, do three things:
1. Ask for references — not a client list, but the chance to speak to a current client. Ask: what happens when something breaks? Do you feel looked after? What annoys you?
2. Ask for a sample monthly report — you will see whether the company produces them at all, and in what form.
3. Ask about their last client incident — how it unfolded, what was done, how long it took. A good partner will tell you straight. A bad one will dodge.
Choosing an IT company is like choosing a mechanic. The cheap “any old one” costs more in lost time than the solid one who charges a fair rate and looks after the car before it breaks down on the motorway.
