The help desk is dead. Long live the service experience.

The ticket queue has had its day

For decades the help desk has worked the same way. Something breaks, a user logs a ticket, the ticket joins a queue, and an engineer eventually picks it up. Success is measured in response times and tickets closed.

That model made sense when work happened in one office on one network. It fits poorly with how organisations operate now. People work from home, the office, client sites and airports. They rely on dozens of cloud applications. They expect technology to behave like the consumer apps on their phones: fast, intuitive and mostly invisible.

When I look at help desk data across different organisations, the same pattern shows up. A large share of requests are repetitive and predictable. Password resets, account unlocks, access requests, new starter setups, printer issues, software installs and "my laptop is slow". These are not complex problems. They are simply problems that the old model makes people wait for.

The cost of that waiting is rarely measured. It is not the engineer's time. It is the sales manager locked out before a pitch, the new starter without a working laptop on day one, and the clinician waiting on an access request. Multiply that across a year and the real cost of a reactive help desk is lost productivity, frustration and, often, workarounds that create security risk.

What a modern service desk looks like

Modernising the help desk is not about replacing people with bots. It is about moving your people up the value chain and letting technology handle the predictable work. In practice, that comes down to five shifts.

From reactive to proactive. Remote monitoring and management platforms such as NinjaOne, alongside Microsoft Intune and Jamf for Apple devices, let us see problems before users do. A disk filling up, a failed patch, a battery about to fail or a device drifting out of compliance can be detected and often fixed automatically. The best ticket is the one the user never has to log.

From manual to automated. Routine tasks such as onboarding and offboarding, licence assignment, group membership, software deployment and patching can be scripted and orchestrated end to end. A new starter's account, mailbox, applications and laptop can be ready before their first morning, and a leaver's access can be removed the moment HR processes the exit. That is a productivity win and a security win.

From queue to self-service. A good self-service portal and knowledge base let people fix common issues in minutes, at any hour. Self-service password reset alone removes one of the most common and frustrating calls to any help desk.

From searching to AI-assisted. AI now earns its place on the service desk. It can triage and categorise tickets, suggest fixes from past resolutions, summarise long email threads and draft knowledge articles. Used well, it helps engineers resolve issues faster and more consistently. Used carelessly, it becomes another source of wrong answers, so we treat it as an assistant to skilled people, not a replacement for them.

From isolated to integrated. The service desk should not sit apart from security. Device health, identity and access, and endpoint protection all feed the same picture. When the service desk and the SOC share context, a suspicious login or an unpatched device is handled as one issue, not two.

What to automate first, and what must stay human

The temptation is to automate everything at once. From my experience, the organisations that get the most value start with a simple filter: automate what is high volume, low complexity and low risk, then expand.

Good first candidates are usually:

• Password resets and account unlocks
• New starter and leaver workflows
• Standard software requests and installs
• Patch deployment and compliance reporting
• Routine device health checks and remediation
• Ticket categorisation and routing

Just as important is knowing what should stay with people. A frustrated executive whose presentation will not project in ten minutes does not want a chatbot. A complex issue spanning the network, a line-of-business application and a vendor needs an experienced engineer who can take ownership. Anything involving sensitive access, a possible security incident or a business-critical outage needs human judgement.

That is why we think of automation as giving time back to engineers, not taking it away. Every hour saved on password resets is an hour available for solving harder problems, improving the environment and talking to the people who use it.

One practical warning: automation built on a messy environment just makes the mess happen faster. Standardise device builds, tidy up identity and access, and document processes before you automate them.

Measure the experience, not just the ticket

If you only measure how fast tickets are closed, you will get a help desk that is very good at closing tickets. A modern service desk should also track:

• Ticket deflection: how many issues were prevented or resolved through automation and self-service
• First contact resolution: how often the problem is fixed the first time
• Repeat incidents: whether root causes are being fixed or just the symptoms
• Onboarding readiness: whether new starters are productive on day one
• User experience: what people actually think of the support they receive
• Device and patch compliance: a direct link between service quality and security posture

These measures tell leadership whether IT is enabling the business, not just keeping up with it.