Why portals, chatbots, and knowledge bases cannot deliver zero tickets
Enterprise IT solved for access when the real problem was resolution
The modern service desk is not short on entry points. In fact, employees can open a portal, search a knowledge base, start a chatbot conversation, or submit a ticket from almost anywhere. In theory, that should mean fewer queues and faster resolution. But if access to IT has improved so dramatically, why has the operational burden behind each interaction barely moved?
That is the tension now facing CIOs and digital workplace leaders. Employees may have more ways to ask for help, but they are still losing time to the same issues, from VPN failures before a customer call to degraded Teams performance during meetings that disrupt productivity and collaboration. Gartner benchmark data shows that self-service resolution remains around 15% of total issues handled, while the average enterprise employee generates roughly eight service desk contacts per year and annual service desk spend per end user has reached $166.
Those numbers point to a difficult reality for IT teams still relying on traditional self-service models. Self-service made it easier to ask for help, but the work of diagnosing and fixing issues still sits elsewhere. Employees still have to interpret guidance, test remediation steps, escalate when workflows fail, and wait for someone with the right context to step in.
A queue-based service desk cannot keep pace as support demand grows, and the risk is slower resolution, rising operational cost, lower employee productivity, and a support model that starts to break under its own volume.
The service desk still absorbs the operational complexity self-service cannot interpret
Traditional self-service is still reactive by design. It improves how employees enter the support process, but it does not remove the process itself. A portal can direct an employee to the right form, a chatbot can summarize a knowledge article, and a virtual assistant can classify intent. But when an issue depends on factors ranging from live device state to user-specific configuration, those systems rarely have enough visibility to resolve the problem themselves.
A Teams issue may appear inside one application, but the cause often sits deeper in the employee’s environment. It may be the device under strain, the network path dropping quality, authentication adding delay, or the active session behaving differently than expected. Many IT issues are far less straightforward than they first appear. What looks simple to the employee is often shaped by conditions they cannot see, and a static workflow cannot interpret. They do not need another article or form. They need the issue diagnosed in context and resolved quickly.
Without live operational visibility, self-service becomes a more polished form of triage. The employee is left to investigate, test fixes, and escalate when the workflow reaches the edge of what it can understand. The ticket has not disappeared. It has just been delayed, with more of the effort pushed onto the employee.
That limitation becomes harder to ignore as enterprise work depends on distributed devices, cloud applications, identity services, and collaboration tools working together throughout the day. When one part of that chain breaks, static self-service is not equipped to understand the full environment well enough to resolve it.
Why now?
The self-service model began when support environments were relatively predictable. Employees worked from managed devices, application portfolios were smaller, and the service desk could absorb the operational complexity that self-service could not.
In 2026, the average environment looks very different. Employees rely on cloud applications, collaboration platforms, identity services, hybrid connectivity, and increasingly AI-powered tools that must function continuously throughout the working day. At the same time, IT leaders are being asked to improve productivity, reduce unnecessary service desk demand, and control cost without lowering the quality of support. As that pressure builds, traditional self-service starts to show its limits. The truth is, it can help employees begin the support process, but it was never designed to understand the full operating context needed to finish it.
The next phase of IT support will be defined by operational execution
The next phase of IT support has to move beyond better access and better conversations. Employees do not judge support by how easily they can start the process. They judge it by whether the problem gets understood, fixed, and removed from their day.
For IT, AI needs to do more than interpret intent or return guidance. It needs enough live context to understand what is happening across the employee’s device, applications, network, and session, along with a safe way to act when the fix is clear. Otherwise, the same work still lands with the service desk, only after another digital handoff.
Zero-friction operations require a different foundation. AI has to reason against current conditions, not just static articles or scripted workflows, and any remediation needs to happen through actions IT has approved, governed, and can improve over time.
Digital employee experience platforms matter in this next model because they give AI the live experience context it needs. Without that context, AI is still trying to support an environment it cannot fully see.
Spark was designed around autonomous resolution rather than ticket deflection
Nexthink Spark was designed for this operating model. Rather than functioning as another chatbot layer in front of an ITSM platform, Spark operates as a personal IT agent on top of Nexthink Infinity’s DEX telemetry, diagnostics, and remediation framework.
When an employee reports an issue through Teams or another conversational interface, Spark can inspect the conditions behind the request while the conversation is still happening. Spark can look at device health, application behavior, network performance, authentication state, and previous remediation activity while the conversation is still happening. It then determines whether there is enough context to resolve the issue through an approved action, or whether it should move to the service desk with the right detail attached.
This transforms the support experience. More issues can be resolved where they occur, rather than being passed into a queue for someone else to investigate later. Employees get back to work faster, IT reduces the everyday friction that quietly slows teams down, and the business gains back time that would otherwise be lost to avoidable digital disruption.
The takeaway for IT leaders
As enterprise environments become more distributed, application-dependent, and operationally complex, IT needs support models that can understand what is happening in the moment and respond appropriately. Resolution increasingly depends on context, not just access to information.
Organizations that continue relying solely on traditional self-service will find it increasingly difficult to balance employee expectations, service desk efficiency, and operational cost.
Nexthink Spark is designed around that shift, combining live DEX telemetry, AI-driven diagnostics, and IT-approved remediation within a personal IT agent that can help resolve issues before they reach the service desk.
For IT leaders, the more important question is what role the service desk should play in the years ahead. If support demand continues to grow while environments become more complex, how much longer can a model built around tickets remain the primary way employees get help?
Get ahead of the change and learn more about Nexthink Spark.
