We are constantly being bombarded with content and hype around the ‘post-covid’ world. Some of this of course is valuable and essential – discussing and planning around how people and organisations should think about the world after lockdown.
In IT this has been particularly life-changing, as we have seen a large amount of change, ‘digital transformation’ and modernisation happen as a necessity rather than a chosen strategy. We have argued about the opportunities and options around ‘digital’ automation, mobile/remote working, culture change, agile ways of working, etc for the last few years, whereas now change has been thrust upon us. In the past, the pros and cons of change have not always been presented, understood, and assimilated as they should. When there is uncertainty, managers will tend to go safe and avoid change. Selling change requires information, clarity, and understanding. People need to feel confident about what they are taking on, before investing and committing to major changes. Much of the mood music and marketing hype around ‘new ways of working’ has confused rather than clarified.
The Covid crisis has moved everything forward quickly and helped to show some of the value of change – moving forward we need to make sure that we still build confidence around new opportunities and not simply drive panic and knee-jerk reactions.
Artificial Intelligence (AI) has been an area where there is now a far greater level of demand and interest – and of course, there are huge potential opportunities and benefits that can be achieved from using this. I have found in recent months that there is increased demand for good quality information around implementation, issues, costs, and benefits – often from those that need to either sell the concept internally or effectively ‘buy’ it – i.e. from change drivers and decision makers who decide on investment and organisational change. The messages need to be simple and clear, not mired in technical detail or supercharged with marketing wow-factor features.
I recently delivered a workshop on *AI for Leaders, with Dan Turchin, on this topic and we realised from that and subsequent contacts that the thirst for information at present has now filtered up to leadership and executive levels, who want simple cost/risk/benefit information to support their decisions. So let’s look at some key issues as a starting point to knowledge and greater confidence around this issue.
AI concerns
There are a number of misconceptions and vagaries that need to be cleared up.
- AI will mean robots taking my job / all of our jobs
- We risk losing human contact by handing this over to machines
- This is expensive
- AI is not proven and can be disastrous
AI clarifications
AI will mean robots taking my job / all of our jobs
AI will be able to replicate work currently done by people, and, in many cases, do it better, more, faster, and cheaper. We shouldn’t stick our heads in the sand about this – there will be challenges to current definitions of jobs and roles. However, the opportunity is that AI will replace a lot of error-prone and repetitive work and thereby provide people with more time and capability to do more difficult and challenging work, and also to think, be creative, and innovative. New technology is always challenging to existing work structures – this has been the case in industrial history for the last 200/300 years. In most cases, jobs may disappear but new work opportunities arise.
IT is also worth bearing in mind, that, in most cases, the uses of AI within an ITSM environment are limited to some very specific functions – e.g. using some RPA (Robotic Process Automation) to ‘unblock’ queues of work or distribute work quickly. In most cases, this represents a small part of the whole end-to-end delivery of a value stream. We are not talking about Ai taking over roles so much as delivering some specific pieces of work that make it work better, faster more consistent etc.
We risk losing human contact by handing this over to machines
Again there is a grain of truth behind this as we will effectively reduce the potential number of human interactions when processes and functions are done via systems. The implications are (1) that we must there ensure that all direct human interactions are of the highest quality, as they will be rare and may have a disproportionate impact on perception, plus (2) we need to manage the transition to automation and review all potential human ‘touchpoints’. A touch point is a direct interaction where value and quality can be improved or reduced, so we need to ensure that these are appropriate to the situation – i.e. it may be that automation is not suitable for some transactions, so these need to be delivered by people.
This is expensive
AI used to be expensive – not any longer. What can take up time and money however is the time and effort required to plan, test, and implement the technology and also manage the transition to it. This is easy to underestimate and many organisations have seen their AI projects drag on in prolonged testing. Any project requires good scoping and business case analysis, so it should be clear what the technology will deliver and what is involved to achieve that.
AI is not proven and can be disastrous
As with any technology, particularly in the hype stage, there are many failures. The technology is constantly evolving of course but the key areas used in e.g. ITSM are established and stable. There is a need to be clear on the business case as mentioned, but also organisational readiness, data readiness (is your CMDB and KB up to date?), process and procedure maturity(can your processes be easily automated? Are they defined and understood?
AI opportunities
So what can we achieve with AI? – here are some basic benefits:
- Optimize the use of available resources – do more with less, get your people to do what you are actually paying them for and not just fixing basic stuff every day
- Improvements in practices and productivity – speed up the flow of standards processes – avoid process jams where work piles up waiting for approval or available resources. Make better use of people’s time
- Reductions in operational costs – less time and resource spent on each task, reduced downtime and cost of failure – speed up delivery with the same resources.
- Improved service levels – offer 24 x 7 availability and the capability to improve turnaround and fulfilment times
- Improvements in customer experience – fast turnarounds and accurate routing of work – avoid errors and offer new options and innovations for service delivery
- Higher availability of business services – automation of monitoring and self-healing sytems can reduce downtime – intelligent systems can also predict and avoid unnecessary failures and associated downtime
- Scalability – less time spent on support and maintenance provides more capability to build new services. Automation can also be built on to develop scale with little impact.
ITSM opportunities
There are many AI options and applications that can deliver value to ITSM practices, such as skills-based routing, machine learning, natural language processing, automated testing, monitoring and deployment, interactive bots and RPA (robotic process automation) – some simple examples:
- Service desk / incident management / request fulfilment – speed up and automate rerouting and escalation of tickets, use RPA to speed up triage and avoid queue jams, automate request approval and fulfilment, automate software deployment
- Knowledge management / problem management – search on unstructured data to provide intuitive knowledge creation, identify trends and patterns not previously seen for problem identification, build on learning and develop new understating of issues
- Info Security / event management / monitoring – predict and avoid failures, take evasive and fast action to resolve issues
- Measurement / reporting / analytics – build and push out real-time information on business impact and performance, provide cumulative analysis of data, and identify business, process, resource and financial trends.
AI represents a number of areas of technology. It is not a ‘one size fits all’ overarching approach and shouldn’t be regarded as such. It is possible to test the water on this with some small restricted trials and application of AI technology – to show benefits and also to identify what is involved with implementation. Whilst the world screams that we should all be using it, it’s important to look at the options and opportunities with clarity and objectivity – to identify the detailed benefits and outcomes that it can be used to deliver for your organisation.
There will doubtless be a number of these that fit your organisation, it’s important to be clear on which of these you choose to aim for. There’s a lot to do but nothing to fear…
* If you are interested in a rerun of this course please contact me at bjr@barclayrae.com.
