There is something wonderfully reassuring about a green dashboard.
Everything is fine. At least, that is what the dashboard would like you to believe. Response time? Green. Resolution time? Green. Availability? Green. SLA attainment? Green.
Meanwhile, somewhere in the organization, an employee has contacted the service desk three times about the same issue, created a workaround involving a spreadsheet that should probably be classified as a controlled substance, and told six coworkers, “Don’t bother opening a ticket. Just call Steve.” Welcome to one of the great contradictions of modern service management.
We have become exceptionally good at measuring service. The harder question is whether we are measuring the things that matter.
I have spent much of my career watching technology transform the service desk — from break/fix and telephone-driven support to enterprise platforms, self-service, automation, virtual agents, analytics and AI. The technology has changed dramatically. Human expectations have changed even faster.
That gap is at the heart of what I call the IT Service Conundrum.
Technology allows us to deliver service faster, cheaper and at greater scale. Touch reminds us that the person consuming that service does not experience a workflow, SLA, knowledge article or automation rule.
They experience whether we made their problem easier or harder to solve. And sometimes, despite all those green indicators, we make it harder.
The Seduction of the Green Dashboard
Dashboards create confidence because they simplify complexity. Leaders can see response times, resolution performance, availability, backlog, satisfaction and provider commitments at a glance.
Red invites attention. Green suggests control.
But a dashboard is not the service. It represents selected aspects of the service based on what we have chosen to measure — and those choices influence behavior. If agents are rewarded primarily for closing tickets, tickets will close. Whether the employee believes the problem was resolved can become secondary. If a provider is measured on call duration, conversations become shorter —even when another two minutes might prevent a second call tomorrow. If self-service deflection is celebrated without measuring customer effort, we may unintentionally reward an experience employees actively dislike.
None of these measures are inherently bad. The problem begins when the measure becomes more important than the outcome it was intended to represent. I have seen organizations with impressive SLA performance and frustrated employees. I have seen technically successful implementations struggle with adoption. I have watched teams improve operational efficiency while the people consuming the service quietly developed workarounds.
That is when service leaders need to ask: Are we measuring the performance of our processes or the effectiveness of our service?
Those are not always the same thing.
From Service Level to Service Experience
For decades, service management has depended heavily on Service Level Agreements. SLAs remain important. Organizations need commitments, providers need accountability and leaders need objective measures. But we should stop expecting SLAs to tell us the entire story. Consider two employees experiencing the same issue. Both contact the service desk. Both receive responses within the SLA. Both tickets are resolved within four hours. The dashboard sees two successful transactions.
Employee A received meaningful communication, understood what was happening and returned to work quickly.
Employee B searched the portal for ten minutes, selected the wrong category because the terminology made sense to IT but not to employees, received an automated message that provided little value, called the service desk for clarification and eventually discovered the ticket had been closed even though additional action was required.
Operationally, both transactions may be green. Experientially, they are completely different.
This is why Experience Level Agreements, or XLAs, are important.
An SLA asks: Did we perform the activity within the agreed target?
An XLA asks: What was the experience of the person receiving the service?
We need both.
Self-Service Should Not Mean “Do It Yourself”
Few areas demonstrate the tech vs. touch problem better than self-service. Organizations understandably want employees to use portals, knowledge bases, virtual agents, automated password resets and digital workflows. These technologies can reduce cost, improve consistency, accelerate fulfillment and free service professionals to concentrate on more complex work. But somewhere along the way, some organizations confused self-service with self-support.
There is a difference. Good self-service removes friction. Bad self-service transfers work from IT to the employee. If I previously called the service desk and explained my problem in two minutes, but the new “efficient” portal requires four menus, twelve fields, internal IT terminology and a guess about which configuration item is affected, we have not automated the service.
We have outsourced part of the service desk to the customer. That may reduce cost on an operational spreadsheet. It may also increase the hidden cost of lost employee productivity.
This is why service leaders should look beyond deflection and measure time to successful outcome, task completion, abandonment, repeat contacts and customer effort. How many clicks does it take to get somewhere useful? How often do employees abandon the process? How many start in self-service and eventually call the service desk anyway? These questions expose something traditional operational metrics often miss: Friction.
And friction is one of the most expensive things we fail to measure.
Automation Is a Design Decision
AI makes this conversation even more important. AI can interpret natural language, summarize incidents, recommend knowledge, categorize requests, identify patterns, generate responses and increasingly take action across enterprise systems. The temptation is obvious: if something can be automated, automate it.
I think that is the wrong starting point.
The better question is: Should this interaction be automated?
In my book, Tech vs. Touch: Navigating the IT Service Conundrum, I use a simple model: Automate. Augment. Engage.
Automate predictable, repeatable transactions where human involvement adds little value —password resets, routine access requests, standard fulfillment, status checks and common questions.
Augment situations where technology can make a human dramatically more effective. Give an agent relevant knowledge before they start searching. Summarize a complicated incident history. Identify similar problems. Recommend the next best action. Remove administrative work so the service professional can concentrate on the customer.
Engage when interactions deserve a human being. A frustrated executive facing a critical business interruption probably does not want a beautifully designed chatbot explaining that it understands their concern. An employee dealing with a complicated onboarding problem may need someone to take ownership across several departments. A major incident may require judgment, communication, reassurance, coordination and leadership.
Technology Changes. Relationships Do Not.
Earlier in my career, I worked in a global oil company supporting a complex environment spanning countries, technologies, vendors and business functions.
There was no shortage of technology. But one lesson stayed with me: service ultimately depended on coordination and trust. Someone had to understand the problem. Someone had to communicate. Someone had to take ownership. And someone had to ensure that multiple teams did not simply complete their individual tasks and declare victory while the business remained unable to work. Those expectations have not disappeared because we now have sophisticated ITSM platforms and AI. If anything, they matter more.
Our systems are getting better at routing work. Automation is getting better at executing tasks. AI is getting better at interpreting requests and recommending actions. That gives us an opportunity to reconsider what human service professionals should be doing. Let technology handle more predictable work. Let people concentrate on ambiguity, judgment, communication, empathy, relationships and business outcomes. That is where human value increases rather than disappears.
Measure What Matters
Mature service organizations need a balanced measurement system. Continue measuring operational health — response, resolution, availability, backlog, throughput and reliability. But add experience — effort, satisfaction, sentiment, communication quality and confidence. Add business outcomes — did we restore productivity? Did the employee accomplish what they were trying to accomplish?
And measure system health — recurring incidents, automation effectiveness, knowledge reuse, failure patterns, technical debt and the conditions creating demand. Together, those measures tell a richer story. Because the objective of service management is not to create green dashboards, but to create effective service.
The Leadership Challenge Ahead
The next chapter of service management will not be defined by whether organizations adopt AI. They will. It will not be defined by whether we automate more transactions. We will.
The real differentiator will be whether leaders understand where technology creates value and where human interaction creates value — and design the service experience accordingly.
We need to resist two extremes: believing every interaction needs a person, and believing every interaction should eventually eliminate one. Somewhere between those positions is a more thoughtful model — one in which technology handles what technology does best and people are given the opportunity to do what people do best. That is the conundrum.
And that is also the opportunity. After more than three decades in IT, I am convinced the future of service management is not really a competition between technology and human touch. It is about learning how to orchestrate both.
So, the next time you are sitting in an operational review and every box on the dashboard is glowing green, enjoy the moment. Then, ask one more question: What are our employees experiencing?
The answer may tell you more about the health of your service organization than the dashboard ever will.
About This Excerpt
This article is adapted from Ric Mims’ upcoming book, Tech vs. Touch: Navigating the IT Service Conundrum, which explores how IT leaders can balance automation, artificial intelligence, service management discipline, employee experience and the human relationships that ultimately determine whether technology delivers value.
Drawing on more than three decades of experience across service management, global IT operations, consulting, organizational change, and technology leadership, the book examines the evolution from the traditional help desk to the AI-augmented digital enterprise — and the leadership decisions required.