Is the concept of the fully automated, AI-driven IT Service Desk a realistic one? And even if it is technically possible, does this make it something that businesses should necessarily aspire to?
Automation and artificial intelligence (AI) are hot topics. Indeed, current estimates suggest that they are set to displace as many as 1 in 5 job roles over the next decade. From customer service and dispatch – right through to high functioning areas such as strategy management, AI has the potential to reshape your organisation from top to bottom.
The IT Service Desk might seem like an obvious candidate for transformation through automation. Sheer workload is perhaps the biggest driver of change: the typical desk finds itself supporting hundreds of systems and applications – along with a rapidly growing number of endpoints. What’s more, with users now utterly reliant on technology to carry out their roles, expectations - especially when it comes to speed of response - have never been higher.
Faced with these pressures over the long term, how do you scale up your capacity? It comes down to two options:
Increase your staffing levels (not a popular option - especially where budgets are tight, market uncertainty is prevalent and where good recruits are in short supply. So this leaves you with the alternative…
Make more efficient use of existing resources. And this is why the promise of automation is such a compelling one. Slash the time you spend on repetitive and routine tasks, streamline your workflow, free up resources, keep control of staffing costs: these potential benefits are hard to ignore.
So what’s the problem?
AI as an inevitability: What’s wrong with this story?
“The more you automate, the better it is for your business”. At least, that’s the impression that you tend to get from discussions around this topic. But just how realistic is this?
Often, businesses are presented with a “roadmap” for transformation: you start by taking sensible steps to automate your workflow - but that’s just the start. You are told that you are on a “journey”; the inevitable conclusion of which is that your service desk is driven almost entirely by AI. After all, who needs real life technicians when you can have a chatbot?
We believe that this narrative deserves interrogation - not least because the needs of businesses vary so widely (a point that is often glossed over whenever this topic is discussed).
Our view is that as with anything else, when it comes to automation, it is possible to have too much of a good thing! Here are the reasons why…
Are tasks suitable for automation?
There is a distinction to be made between AI and intelligent automation. AI utopia would be training up a robot, like you would a human, to be a Service Desk Analyst. We’re a long way from that as a reality. However, intelligent automation is more feasible, taking tasks of the job function and automating parts of it.
The prime targets for intelligent automation are those tasks that are easily repeatable. Take a chatbot, for instance. This type of technology can be extremely useful for helping customers find fixes to commonly encountered problems. As such, they can have a big role to play in freeing up help desk resources. Not least, they mitigate the need for desk staff to respond to the same, routine issues again and again.
All of this is fine if your service desk encounters relatively large quantities of the same types of issues. The trouble is, this is not always the case for many organisations. The average enterprise has close to 500 apps and systems in use. Some common issues might crop up frequently, but by no means all (or even, a significant proportion of them).
Automation & AI start to lose their value where the issues encountered tend to be of a complex nature. So if the proportion of commonly recurring tasks encountered by your team is relatively low, there’s a very real danger that the benefits realised through automation will be outweighed by the costs and effort of implementation.
Machine learning may be clever – but that’s not to say that you can let the machines get on with it without any human intervention.
Business IT infrastructure rarely stands still. Systems are regularly tweaked; new software is introduced and processes are subject to change. For it to work properly, you need to make sure that your AI platform has access to all relevant information – otherwise the results it produces could fall way short of a useful solution. ‘Feeding’ your AI bot with updated information – and testing those bots before the changes are rolled out – can both be time consuming processes.
The value of the human touch
A big promise of AI is that it will increase customer satisfaction. It enables uncomplicated issues to be resolved with little or no human contact and in the quickest possible time, freeing up technicians to devote more time to the most significant events.
But just be aware that even for relatively minor issues, speed of response is not the sole driver user satisfaction.
Regardless of the severity of the issue, Accenture found that 73% of people prefer human interactions where a problem needs to be fixed or where advice is needed. People trust other people over and above machine-driven platforms – no matter how smart that platform might be.
The extent to which you take into account misgivings over the absence of human interaction can depend on where the pushback is coming from. It may make business sense to require ‘low value’ users to accommodate a chatbox solution (especially for routine queries)- even though the slightly impersonal element may take some getting used to initially. Pushback could be more of an issue for users higher up the corporate food chain; for whom having to grapple with automated responses may be (in their view) out of the question – especially if they have a business critical deal to push through in the next five minutes.
Alongside this, it’s also the case that an AI platform is only ever as good as its programmer. If that platform is perceived as capable of providing only generic or semi-relevant solutions to individual problems, loss of trust will be an inevitable consequence.
When it comes to your service desk, confidence is everything – otherwise users could be tempted to bypass the desk completely. If this confidence is eroded through the imposition of poorly conceived or implemented AI solutions, it can be very difficult to win it back.
Automation is not a case of ‘all or nothing’
How can we make this better? For each process your service desk is responsible for, this is the key question to ask. With this in mind, it’s certainly the case that automation has a big role to play. Here are just a few examples:
It can help increase user and Service Desk productivity, reducing the need for human intervention and enabling self-service for routine requests and issues.
There is definitely an opportunity to unlock the capability already available within existing toolset software to increase the level of automation in processes. For instance in aiding with ticket categorisation and triage, re-routing those tickets to appropriate teams automatically.
Automation can streamline communications through configuring automatic updates and alerts.
There’s potential to increase automation in the routine checking processes many Service Desk analysts have to do – for instance to check for PC updates etc.. We see an increasing of specialist software being developed and now available to do this.
Intelligent analytics is a big area where actually a robot may be better than a human at identifying trends from big data and giving useful insights into performance and showing you how and why bottlenecks occur.
Perhaps we should stop talking about automation as a ‘journey’ with a workforce of robots as the inevitable endpoint. Instead, businesses and Service Desk management, should focus on the real life benefits intelligent automation brings on a process by process level through continuous improvement. However, it’s also worth remembering and leveraging the power only humans can bring such as empathy, ideas and judgement where appropriate on your service.
This way, your Service Desk and IT department will be able to realise the full benefits of all types of resources available to them.