It's nine in the morning. An employee tries to open the VPN and it won't connect. They can't start work until it does. They open a ticket, describe the problem and hit send. The ticket goes into the queue. Sometime in the afternoon, someone gets to it. Until then, the work sits idle.
This is the service desk's basic promise: when something breaks, you open a ticket and wait. It was a sensible answer back in the 2010s, when there was no alternative. Now there is.
What the ticket queue really costs
The cost of a service desk isn't the monthly fee you see on the invoice. It's the waiting.
When an employee is stuck and waits two hours for an answer to a simple problem, the company pays two hours of salary for work that doesn't get done. Multiply that by the number of tickets per month, and next to the monthly fee appears a second, larger number that nobody invoices separately. It shows up nowhere, but it's real.
Most tickets are the same routine, over and over. A password needs resetting. Software needs installing. The VPN won't connect. The printer can't be found. These don't need an expert. They need an answer that already exists somewhere, usually inside someone else's head or in a document nobody can find.
What replaces the routine
Two things together: AI and documented processes. Neither is enough on its own.
When a company's procedures, settings and most common fixes are written out clearly, AI can answer from them instantly. The employee asks in plain language, "the VPN won't connect, what do I do," and gets an answer in seconds — not in a queue, not two hours later. My estimate from practical work is that this handles the bulk of what a ticket is opened for today.
Examples of what gets resolved without a ticket:
- Password reset and locked accounts. Guided self-service, no queue.
- Software installation. Pre-packaged, the employee installs it themselves following a clear instruction.
- VPN and connection issues. Step-by-step troubleshooting, the same questions a support person would ask.
- "How do I do this" questions. Email rules, shared folders, meeting settings. The answer is in a guide, not a queue.
This isn't utopia. It's documentation plus a properly built environment for the AI to work on top of. The biggest job isn't the AI. It's writing the processes out, once, properly.
A decision-maker's calculation
Work out your own hidden cost. You need three figures you either know or can estimate:
- Tickets per month (how many support requests your team makes)
- Average wait time per ticket in hours (queue time, not time spent solving)
- Employee cost per hour (salary including overheads)
Tickets × wait time × hourly cost = monthly idle time in euros. This sum doesn't appear on the support contract invoice. Compare it against getting the same answer in seconds.
AI doesn't replace the expert
This is where many decision-makers tense up: are people being replaced by a machine? No. AI replaces the ticket system, not the expert.
The routine is the part worth moving to the machine. But someone builds the environment the machine works in: writes the processes out, defines what the AI may and may not answer, handles the architecture, the security and whatever goes wrong. That's the expert's work, and its importance grows rather than shrinks. The difference is that the expert no longer spends their time resetting passwords.
This follows directly from how I think about an IT environment: build one that fixes the routine itself, and leave the human what actually needs judgement. (More on this: the self-healing IT environment.)
Where to start
Not by swapping the service desk for AI overnight. First you look at what your tickets actually consist of. Most are probably the same repeating routine, and that's exactly the part worth handling differently.
Want to know how much of your support load is routine that AI and proper documentation would handle? You find that out in an IT Health Check, where I go through your environment and tell you concretely what AI can handle and what needs a human. If you want a continuous grip on it, an ongoing IT partnership keeps the environment in shape and evolves with you.