When a “great” service desk is a warning sign
By Mark Boyer, Service Operations Director, Softcat
We have a strange habit in IT. When the system is broken, we do not redesign the system. We celebrate the people who cope with it.
We hand out praise for resilience. We reward heroics. We tell stories about someone who stayed late, pulled a rabbit out of a hat, and saved the day. And to be clear, I love those people. Service desk professionals are often the most emotionally intelligent, commercially aware, and customer-minded people in our industry.
There is an uncomfortable truth. A consistently “great” service desk can be evidence that the organisation has normalised poor service design elsewhere. It becomes a high-performing shock absorber for problems that should never have reached it in the first place. Now that does not make the desk bad; it makes the desk a signal.
The shift we need to admit
If you work on a service desk, you will recognise this pattern instantly. The job is meant to be simple: listen, understand, solve, restore. However, over time, the work becomes something else. It becomes translation, mediation, navigation, and apology.
You are not just resolving issues. You are helping people navigate a maze of teams, tools, ownership boundaries, and processes designed for internal convenience rather than human experience. The service desk becomes the place customers come to “get things done” because the organisation has not designed a clear path for them to do it any other way.
And when that happens, we start improving the wrong thing. We optimise the desk’s handling of demand rather than questioning why it exists at that volume and shape in the first place.
Stop asking people to be exceptional to compensate for the system
There is a line I have started using that tends to split a room: if your service desk has to be great to keep the organisation running, it is evidence that the service itself is not.
That is not meant to be a cheap slogan. There is a practical point behind it. When the service is poorly designed, the desk becomes the compensator; it absorbs ambiguity, friction, missing context, badly designed change, and recurring failure patterns. It is not uncommon for desks to become brilliant at extracting the real issue from vague descriptions, triaging based on intuition when the data is incomplete, bending the process because it has become a substitute for thinking, and maintaining customer confidence even when the system does not deserve it.
That is craft, that is professionalism, that is human brilliance. It is also a sign that the organisation is relying on human brilliance as a strategy. It is not scalable, it is not fair, and it creates a hidden tax on both customers and colleagues.
Three practical shifts that start the change
This is usually the point where organisations reach for a framework, a new tool, or a maturity model. I would argue you do not need any of that to begin shifting the system. You just need three decisions.
First, treat demand as a design problem, not an operational burden.
The volume of tickets you receive is not just “work”; it is feedback. If the same issues keep coming up, your service is telling you something. Not “train users better” and not “write another knowledge article”. Something simpler: the service is not designed for real human behaviour. Start by logging the top reasons people contact the desk in plain English, without taxonomies, categories, or internal language. Then ask one design question: what would need to be true for this contact not to happen at all? Not resolved faster, not handled better. Not happen. That alone shifts the conversation from coping to redesigning.
Second, refuse speed as the definition of success.
Speed is seductive because it is measurable, it also creates behaviour you do not want. If you reward time to close, you encourage shallow fixes and premature closure. If you reward first-contact resolution blindly, you encourage the avoidance of complexity. If you reward handling time, you encourage deflection rather than understanding. Metrics get gamed, not because people are malicious, but because incentives shape behaviour. The alternative is “no metrics”, it is a better measure. Measure fewer things but make them matter. Reduction in repeat demand and handoffs. Confidence restored, not ticket closed. Customer effort reduced, not customer satisfaction surveyed. That is how you move from performance theatre to genuine service improvement.
Third, protect judgment and autonomy as part of your operating model.
The service desk is full of human judgment because it has to be. People call when the world is messy, when they are stuck, when the workflow is broken, when the context is missing. Yet many organisations respond to mess by removing autonomy. More scripts, more forms, more mandatory fields, more routing rules. This is how you create a support function that is compliant but ineffective. The best desks operate like a good sports team: small enough to hold each other to a standard, clear on the objective, and trusted to choose the path. Morale becomes a multiplier. If you want better outcomes, protect the conditions that allow judgment to thrive.
Why this matters for service desk and ITSM teams
If you run a service desk today, you sit in one of the most valuable positions in IT. You are closest to friction, closest to customer effort, closest to the truth of how services are experienced. That means you can do more than improve response times; you can shape what the organisation becomes.
But it also means you are at risk of being used as a buffer. If your desk is celebrated as “great” while the same demand patterns persist, you are not being valued; you are being relied upon. And that reliance is often invisible to leadership because the desk is working. Tickets are flowing, SLAs are green. This is why great desks can be dangerous; they hide the redesign work that needs to happen upstream.
Closing thoughts
There is a future where support becomes more preventative, more embedded, and more human. Not because we declare ourselves “AI native” and buy a feature. AI is far from established enough for that kind of certainty.
The real shift is simpler. Stop designing your service organisation around how work is routed. Start designing it around how work is experienced. Stop rewarding speed. Start rewarding reduced friction. Stop asking for heroics. Start improving the system so heroics are no longer required.
If you are attending SDI Spark26, I will go deeper into the provocative edge of this argument and the practical shape of what comes next.
But you do not need to wait for a conference to start. If your service desk has to be great to keep the organisation running, the most valuable work you can do is not to make the desk even better. It is to redesign the service so that it does not need saving.
