Service desk is the cleanest paid on-ramp into UK infrastructure work, and the seat most people sit in eighteen months too long. The job will teach you more in your first year than any course ever will. The risk is that after about two years the queue stops teaching you and starts paying you to coast, and most people don't notice the day that changes.
Who this is for
- You've got no commercial IT on your CV and need a paid seat where someone will actually train you.
- You can stomach a ticket queue full of password resets, printer drivers and broken Outlook profiles without losing your mind.
- You're happy to be the most junior person in the room for twelve to eighteen months.
- You'll keep tinkering outside work, because the queue alone will not grow you past second-line.
Who this isn't for
- You're already running a small sysadmin or netadmin patch somewhere. Going back to first-line is a paid sideways step.
- You need fully remote from day one. Most decent UK first jobs are hybrid, often three days on site near a server cupboard.
- You want a clean engineering problem to chew on. The queue is interruption-driven, not project-driven.
- You're banking on the role itself promoting you. It won't. The seat is a launchpad, not an escalator.
What people believe
The standard belief is that service desk is the foot in the door, and that if you're competent, work hard, close tickets, and pass a few certs, the team will naturally move you into sysadmin, then infrastructure, then cloud or security. Stay long enough and the next role finds you.
What actually happens
What actually happens is that the queue is designed to consume the people who are good at it. The better you are at closing tickets, the more tickets you get. Promotions go to the person who steps outside the queue and starts doing work the team needs but nobody has been hired to do: a knackered Active Directory cleanup, a VPN rebuild, the Linux box nobody admits to owning. People who quietly do the queue at 95% for three years almost never make sysadmin internally. The ones who make it move within eighteen to twenty-four months, and usually by jumping company, not by waiting.
Why the gap exists
Service desk teams are cost centres. The business measures them on ticket volume, response time and customer satisfaction, not on the technical growth of the people in them. There's no structural reason for your manager to push you upwards. The pay is enough to live on, the work is varied enough to feel like you're learning, and the team is usually friendly. Every incentive points at staying. The only person watching whether you're still growing is you, and after the first eighteen months that gets quieter every quarter.
Why people get stuck
Around the eighteen-month mark the queue stops being scary. You know the systems, the team likes you, the pay covers the rent, and the work has stopped surprising you. That feels like competence. It's actually the moment the job stops teaching you anything new. People who notice it within six months tend to move. People who don't are still on the desk five years later wondering why nobody promoted them.
You pass CompTIA Network+, then Security+, then start on AZ-900, and it feels like progress. It is, slightly. But certs on their own won't move you off the desk. Hiring managers for the next role want to see you've already done some of the work, even badly, on a real network or in a real tenancy. Stacking certs without ever touching the kit they describe is the most common way people spend two years preparing for a jump that never happens.
The single most common stall on the desk is assuming that doing the job well is a strategy. It isn't. The people who get out wrote the runbook nobody asked for, fixed the AD problem the sysadmins kept dodging, or quietly took on the user-creation script and turned it into something the team actually uses. Visibility on a service desk doesn't come from ticket counts. It comes from doing one piece of work a quarter that the queue didn't generate.
What to do about it
- Pick a date eighteen months from your start and write it down. That's your review date, not your leaving date, but treat it as real.
- Inside the first six months, find the one system nobody on the desk wants to touch. Active Directory, Group Policy, the wifi controller, the print server. Become the person who fixes it.
- Keep a running list of every ticket type that taught you something genuinely new this month. When that list drops below five for two months running, your queue has stopped growing you.
- Before you sit your second cert, deploy the thing the cert describes somewhere a hiring manager could see. A home lab, a free Azure tenancy, a GitHub repo with terraform that builds it. The cert and the artefact together are worth far more than the cert alone.
- When you apply out, apply for the role one step ahead of where you currently sit, not two. First-line to second-line, second-line to junior sysadmin or junior cloud. Trying to skip a rung from the desk is the most common reason good candidates get screened out before interview.
Named failure mode
The defining failure mode of service desk careers is the person who becomes structurally too useful to lose. They know every system, every shortcut, every user's name. The team relies on them, the business pays them a little more each year to stay, and nobody, including them, has any incentive to move them on. Five years in they're earning thirty-two grand, doing the same job they did at year two, and the market has decided they're a service desk lifer. The exit gets harder every year they stay.
What would change this verdict
- Your manager hands you ownership of a specific platform (AD, a backup system, an MDM tenancy) with your name on the documentation. That changes the verdict because you're no longer in the queue, you're a junior admin with a desk job on the side.
- The team restructures and second-line is merged into infrastructure. If your role moves with it, the launchpad timer resets and the seat becomes worth holding for another twelve to eighteen months.
- You're hired into a service desk that's explicitly a rotation programme (some MSPs and a handful of large enterprises run these) with a written commitment to move you into a defined team within twenty-four months. That's a different job with the same title.
- You're using the desk to fund a deliberate cross-pivot (data, software, security) and you've already started building outside it. In that case the seat is a salary, not a career, and the verdict stops applying.
What POST has observed
People who leave the desk successfully tend to move to a junior sysadmin or junior cloud role at a different company on roughly the same money, sometimes a little less. The pay rise comes at the second move, not the first. People who hold out for a promotion that doubles their salary in one go usually stay on the desk.
Two years on an MSP desk tends to produce a more employable engineer than four years on an in-house desk, because the exposure to varied estates is wider. The cost is sharper. Burnout, on-call, weekend work and a customer base that treats you as disposable. Worth knowing before you take the role, not after.
Where it exists, the strongest pattern is: service desk for twelve to eighteen months, then a sideways move into a small infrastructure or platform team in the same company, then external move at the two to three year mark. The internal sideways step is the bit most people skip, and it's the bit that does the most work.